{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# pandas - data manipulation and analysis with dataframes\n", "\n", "[pandas.pydata.org](https://pandas.pydata.org/)\n", "\n", "Pandas is a \"high-level\" module which depends heavily on the \"low-level\" numpy package. Pandas is more friendly for statistics/exploratory analysis.\n", "Like numpy or matplotlib, it is part of the scipy project.\n", "\n", "A great strenght of pandas is its **DataFrame** which emulates many of the convenient behavior and syntax of their eponym counterpart in the **R** language.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's begin by importing pandas:" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import pandas as pd" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## The dataset" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To illustrate pandas functionalities, we will use a file that contains gene expression data. \n", "\n", "The study from which the data came investigated the stress response in the hearts mice deficient in the SRC-2 gene (transcriptional regulator steroid receptor coactivator-2). The data can be found here: http://www.ncbi.nlm.nih.gov/pubmed/23300926.\n", "\n", "The data are structured as follows:\n", "\n", "* each row contains the expression values of a particular gene\n", "* each column corresponds to one sample/condition and contains the expression of values of all genes in that sample\n", "\n", "Take a look at the file first in a text editor or on the command line to verify its structure.\n", "\n", "The sample names are given in the first row (header). \n", "\n", "Based on the names, we can guess that we have gene expression values for heart tissue of two types: \"WT\" (wildtype) and \"KO\" (knock out), and four replicates for each condition:\n", "\n", "Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 Heart_KO_2 Heart_KO_3 Heart_KO_4" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 1. Reading data from a file into a dataframe" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We first use the function [read_csv](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.read_csv.html) to read the file. By default, as the name suggests, this function looks for csv files. \n", "\n", "> In case of big datasets it is convenient to look at a fraction of the data. For this the functions [head](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.head.html) or [tail](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.head.html) are helpful. " ] }, { "cell_type": "code", "execution_count": 106, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1\\tHeart_WT_2\\tHeart_WT_3\\tHeart_WT_4\\tHeart_KO_1\\tHeart_KO_2\\tHeart_KO_3\\tHeart_KO_4
01415670_at\\t1214.447\\t1182.464\\t1206.226\\t1196...
11415671_at\\t3490.098\\t2882.784\\t2650.033\\t2934...
21415672_at\\t4510.369\\t4292.057\\t4071.057\\t4275...
31415673_at\\t598.8334\\t385.0178\\t458.4872\\t514....
41415674_a_at\\t1400.325\\t1328.295\\t1416.923\\t13...
\n", "
" ], "text/plain": [ " Heart_WT_1\\tHeart_WT_2\\tHeart_WT_3\\tHeart_WT_4\\tHeart_KO_1\\tHeart_KO_2\\tHeart_KO_3\\tHeart_KO_4\n", "0 1415670_at\\t1214.447\\t1182.464\\t1206.226\\t1196... \n", "1 1415671_at\\t3490.098\\t2882.784\\t2650.033\\t2934... \n", "2 1415672_at\\t4510.369\\t4292.057\\t4071.057\\t4275... \n", "3 1415673_at\\t598.8334\\t385.0178\\t458.4872\\t514.... \n", "4 1415674_a_at\\t1400.325\\t1328.295\\t1416.923\\t13... " ] }, "execution_count": 106, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv(\"GSE41558_series_matrix.tsv\")\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Take a look at how the data have been read: since read_csv() did not find any comma in the file, it treated each entire line as one entry. To correctly read this file, we have to specify that the column separator is a tab:" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv(\"GSE41558_series_matrix.tsv\", sep='\\t')\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's check the dimensions of the dataframe:" ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(45101, 8)" ] }, "execution_count": 33, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.shape" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can also check the column names:" ] }, { "cell_type": "code", "execution_count": 34, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Index(['Heart_WT_1', 'Heart_WT_2', 'Heart_WT_3', 'Heart_WT_4', 'Heart_KO_1',\n", " 'Heart_KO_2', 'Heart_KO_3', 'Heart_KO_4'],\n", " dtype='object')" ] }, "execution_count": 34, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.columns" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now it is obvious that we read the dataframe properly. \n", "\n", "Note that pandas made an additional interpretation. \n", "\n", "Namely, that the first column (containing gene names)is not a normal column but corresponds to the lines indexes (row labels) of the dataframe. \n", "\n", "The reason the gene names were used as index is because the first line in the `GSE41558_series_matrix.tsv` file contained 8 entries and all the other lines contain 9 entries. So pandas automatically inferred that the first column should be the index." ] }, { "cell_type": "code", "execution_count": 35, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Index(['1415670_at', '1415671_at', '1415672_at', '1415673_at', '1415674_a_at',\n", " '1415675_at', '1415676_a_at', '1415677_at', '1415678_at', '1415679_at',\n", " ...\n", " 'AFFX-r2-P1-cre-5_at', 'AFFX-ThrX-3_at', 'AFFX-ThrX-5_at',\n", " 'AFFX-ThrX-M_at', 'AFFX-TransRecMur/X57349_3_at',\n", " 'AFFX-TransRecMur/X57349_5_at', 'AFFX-TransRecMur/X57349_M_at',\n", " 'AFFX-TrpnX-3_at', 'AFFX-TrpnX-5_at', 'AFFX-TrpnX-M_at'],\n", " dtype='object', name='gene', length=45101)" ] }, "execution_count": 35, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.index" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To better understand how pandas reads in data let's use the same file, after stripping the header line. " ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
1415670_at1214.4471182.4641206.2261196.031174.6181184.458985.55031214.54
01415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
11415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
21415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
31415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
41415675_at497.7774441.8907463.8198471.8565404.6885357.1436442.7815519.2141
\n", "
" ], "text/plain": [ " 1415670_at 1214.447 1182.464 1206.226 1196.03 1174.618 \\\n", "0 1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1 1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "2 1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "3 1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "4 1415675_at 497.7774 441.8907 463.8198 471.8565 404.6885 \n", "\n", " 1184.458 985.5503 1214.54 \n", "0 2823.2600 2721.8840 2790.8340 \n", "1 4045.9900 4553.7360 4358.6350 \n", "2 544.3807 569.1154 323.8668 \n", "3 1462.9640 1237.2440 1797.9060 \n", "4 357.1436 442.7815 519.2141 " ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv(\"GSE41558_series_matrix_no_header.tsv\", sep='\\t')\n", "df.head()\n" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(45100, 9)\n" ] } ], "source": [ "print(df.shape)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now pandas identified 9 columns, but it used the first row to get column names. \n", "\n", "To prevent this, we need to tell pandas explicitly that there is no header in the file." ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
012345678
01415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
11415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
21415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
31415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
41415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " 0 1 2 3 4 5 \\\n", "0 1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1 1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "2 1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "3 1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "4 1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " 6 7 8 \n", "0 1184.4580 985.5503 1214.5400 \n", "1 2823.2600 2721.8840 2790.8340 \n", "2 4045.9900 4553.7360 4358.6350 \n", "3 544.3807 569.1154 323.8668 \n", "4 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv(\"GSE41558_series_matrix_no_header.tsv\", sep='\\t', header=None)\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This looks much better as there is no misinterpretation of the actual data. \n", "\n", "Note that by default pandas numbers the columns (and rows), starting from index 0." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "One advantage of dataframes over other data structures is that it allows one to combine different data types in a single data structure. We can therefore keep track and index/slice by row and column names." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 1.1 Setting column names" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's say we have a file that only contains values and we want to attach the column labels. First make the list of column names:" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "column_names = [\"gene\", \"Heart_WT_1\", \"Heart_WT_2\", \"Heart_WT_3\", \"Heart_WT_4\", \"Heart_KO_1\", \"Heart_KO_2\", \"Heart_KO_3\", \"Heart_KO_4\"]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "and then attach it to the data frame: " ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
geneHeart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
01415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
11415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
21415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
31415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
41415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " gene Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "0 1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1 1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "2 1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "3 1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "4 1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "0 1184.4580 985.5503 1214.5400 \n", "1 2823.2600 2721.8840 2790.8340 \n", "2 4045.9900 4553.7360 4358.6350 \n", "3 544.3807 569.1154 323.8668 \n", "4 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.columns = column_names\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 1.2 Setting the index" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Note that above, gene names are just another column (first) of the data frame, and that there is an additional index (numeric) of the rows. However, since in this example the gene names are unique it is much more convenient to use them as index, which we would do as follows (modifying the dataframe 'in place'):" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.set_index('gene', inplace=True) \n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 2. Slicing dataframes" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's say we want to select part of a dataframe. \n", "This could be done either with the [`.iloc`](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.iloc.html) or the [`.loc`](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.loc.html) methods.\n", "\n", " * `.loc` is used to select rows and column using their label (*e.g.* gene names in our case)\n", " * `.iloc` is used to select rows and column using their position (*i.e.* from 0 to the number of rows/columns - 1 )\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For example, to get the first three rows of the dataframe we could do (recall that the interval is open on the right side):" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.4471182.4641206.2261196.0301174.6181184.458985.55031214.540
1415671_at3490.0982882.7842650.0332934.8612723.9762823.2602721.88402790.834
1415672_at4510.3694292.0574071.0574275.2764127.9414045.9904553.73604358.635
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226 1196.030 1174.618 \n", "1415671_at 3490.098 2882.784 2650.033 2934.861 2723.976 \n", "1415672_at 4510.369 4292.057 4071.057 4275.276 4127.941 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.458 985.5503 1214.540 \n", "1415671_at 2823.260 2721.8840 2790.834 \n", "1415672_at 4045.990 4553.7360 4358.635 " ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.iloc[0:3]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or the first three columns:" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3
gene
1415670_at1214.44701182.46401206.2260
1415671_at3490.09802882.78402650.0330
1415672_at4510.36904292.05704071.0570
1415673_at598.8334385.0178458.4872
1415674_a_at1400.32501328.29501416.9230
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260\n", "1415671_at 3490.0980 2882.7840 2650.0330\n", "1415672_at 4510.3690 4292.0570 4071.0570\n", "1415673_at 598.8334 385.0178 458.4872\n", "1415674_a_at 1400.3250 1328.2950 1416.9230" ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.iloc[:,0:3].head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or we can select both a subset of the rows and the columns:" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3
gene
1415670_at1214.4471182.4641206.226
1415671_at3490.0982882.7842650.033
1415672_at4510.3694292.0574071.057
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226\n", "1415671_at 3490.098 2882.784 2650.033\n", "1415672_at 4510.369 4292.057 4071.057" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.iloc[0:3,0:3]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Using the column or the row names for indexing :" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3
gene
1415670_at1214.4471182.4641206.226
1415671_at3490.0982882.7842650.033
1415672_at4510.3694292.0574071.057
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226\n", "1415671_at 3490.098 2882.784 2650.033\n", "1415672_at 4510.369 4292.057 4071.057" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.loc[ ['1415670_at','1415671_at','1415672_at'],['Heart_WT_1', 'Heart_WT_2', 'Heart_WT_3']]" ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3
gene
1415670_at1214.4471182.4641206.226
1415671_at3490.0982882.7842650.033
1415672_at4510.3694292.0574071.057
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226\n", "1415671_at 3490.098 2882.784 2650.033\n", "1415672_at 4510.369 4292.057 4071.057" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "#or \n", "df.loc[ '1415670_at':'1415672_at' , 'Heart_WT_1':'Heart_WT_3' ]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "**Important note**\n", "\n", "If you want to mix accession by position and labels at the same time, you may do so by combining `.loc` with `.index` or `.columns` \n", "\n", "\n", "\n", "> [there exists a method called `.ix` but it is currently deprecated](https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#ix-indexer-is-deprecated)" ] }, { "cell_type": "code", "execution_count": 31, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3
gene
1415670_at1214.4471182.4641206.226
1415671_at3490.0982882.7842650.033
1415672_at4510.3694292.0574071.057
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226\n", "1415671_at 3490.098 2882.784 2650.033\n", "1415672_at 4510.369 4292.057 4071.057" ] }, "execution_count": 31, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.loc[ df.index[0:3] , 'Heart_WT_1':'Heart_WT_3' ]" ] }, { "cell_type": "code", "execution_count": 32, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3
gene
1415670_at1214.4471182.4641206.226
1415671_at3490.0982882.7842650.033
1415672_at4510.3694292.0574071.057
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226\n", "1415671_at 3490.098 2882.784 2650.033\n", "1415672_at 4510.369 4292.057 4071.057" ] }, "execution_count": 32, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.loc[ '1415670_at':'1415672_at' , df.columns[0:3] ]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 3. Selecting/Filtering dataframes" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Lets read the dataframe again." ] }, { "cell_type": "code", "execution_count": 36, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 36, "metadata": {}, "output_type": "execute_result" } ], "source": [ "column_names = [\"gene\", \"Heart_WT_1\", \"Heart_WT_2\", \"Heart_WT_3\", \"Heart_WT_4\", \"Heart_KO_1\", \"Heart_KO_2\", \"Heart_KO_3\", \"Heart_KO_4\"]\n", "df = pd.read_csv(\"GSE41558_series_matrix_no_header.tsv\", sep='\\t', header=None)\n", "df.columns = column_names\n", "df.set_index('gene', inplace=True) \n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "And filter it based on some criteria that we may be interested in. \n", "\n", "For example say we want to find genes that have at least 250 reads in the 'Heart_WT_1' sample. We would do it like this: first we find the genes that satisfy the condition:" ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "myslice = df['Heart_WT_1']>250\n", "print(type(myslice))" ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "gene\n", "1415670_at True\n", "1415671_at True\n", "1415672_at True\n", "1415673_at True\n", "1415674_a_at True\n", "Name: Heart_WT_1, dtype: bool" ] }, "execution_count": 38, "metadata": {}, "output_type": "execute_result" } ], "source": [ "myslice.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Applying the `>` operator returns a boolean Series with the result of the function on every element of the Series. Then, to select the corresponding elements of the dataframe, we use the boolean Series to slice the original dataframe:" ] }, { "cell_type": "code", "execution_count": 39, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "mymysteriousobj = df[df['Heart_WT_1']>250]\n", "print(type(mymysteriousobj))" ] }, { "cell_type": "code", "execution_count": 40, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 40, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[df['Heart_WT_1']>250].head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can design more complicated filters, as below, we select genes that have more than 250 reads in WT samples, less than 150 in all KO samples:" ] }, { "cell_type": "code", "execution_count": 41, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1429553_at399.4376447.415674.9503320.7553117.62532.0594711.2015937.07462
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1429553_at 399.4376 447.415 674.9503 320.7553 117.625 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1429553_at 32.05947 11.20159 37.07462 " ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[(df['Heart_WT_1']>250) &\n", " (df['Heart_WT_2']>250) &\n", " (df['Heart_WT_3']>250) &\n", " (df['Heart_WT_4']>250) &\n", " (df['Heart_KO_1']<150) &\n", " (df['Heart_KO_2']<150) &\n", " (df['Heart_KO_3']<150) &\n", " (df['Heart_KO_4']<150)]" ] }, { "cell_type": "markdown", "metadata": { "collapsed": true }, "source": [ "We can also slice the result of filtering. For example, let's say that we want to extract the genes with more than 250 reads in the first WT and less than 50 reads the first KO sample but then also only keep these two columns of the data." ] }, { "cell_type": "code", "execution_count": 44, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_KO_1
gene
1416649_at252.742342.268160
1416677_at419.528048.156210
1416913_at276.170312.551740
1417246_at416.631429.861870
1417600_at490.809540.921280
1419358_at281.561720.607560
1420926_at281.353032.450370
1432499_a_at319.433013.148060
1434442_at443.65343.316613
1435219_x_at293.76484.952919
1440250_at261.224442.827180
1440823_x_at256.497244.901100
1449509_at339.966419.187620
1456471_x_at305.597921.563080
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_KO_1\n", "gene \n", "1416649_at 252.7423 42.268160\n", "1416677_at 419.5280 48.156210\n", "1416913_at 276.1703 12.551740\n", "1417246_at 416.6314 29.861870\n", "1417600_at 490.8095 40.921280\n", "1419358_at 281.5617 20.607560\n", "1420926_at 281.3530 32.450370\n", "1432499_a_at 319.4330 13.148060\n", "1434442_at 443.6534 3.316613\n", "1435219_x_at 293.7648 4.952919\n", "1440250_at 261.2244 42.827180\n", "1440823_x_at 256.4972 44.901100\n", "1449509_at 339.9664 19.187620\n", "1456471_x_at 305.5979 21.563080" ] }, "execution_count": 44, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[(df['Heart_WT_1']>250) & (df['Heart_KO_1']<50)][ ['Heart_WT_1', 'Heart_KO_1'] ]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can very easily keep just the names of these genes in a list:" ] }, { "cell_type": "code", "execution_count": 45, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "['1416649_at', '1416677_at', '1416913_at', '1417246_at', '1417600_at', '1419358_at', '1420926_at', '1432499_a_at', '1434442_at', '1435219_x_at', '1440250_at', '1440823_x_at', '1449509_at', '1456471_x_at']\n" ] } ], "source": [ "list_of_genes = list(df[(df['Heart_WT_1']>250) & (df['Heart_KO_1']<50)].index)\n", "print(list_of_genes)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "A very powerful feature is that we can apply filters on all columns and rows at the same time:" ] }, { "cell_type": "code", "execution_count": 50, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_atTrueTrueTrueTrueTrueTrueFalseTrue
1415671_atTrueTrueTrueTrueTrueTrueTrueTrue
1415672_atTrueTrueTrueTrueTrueTrueTrueTrue
1415673_atFalseFalseFalseFalseFalseFalseFalseFalse
1415674_a_atTrueTrueTrueTrueTrueTrueTrueTrue
...........................
AFFX-TransRecMur/X57349_5_atFalseFalseFalseFalseFalseFalseFalseFalse
AFFX-TransRecMur/X57349_M_atFalseFalseFalseFalseFalseFalseFalseFalse
AFFX-TrpnX-3_atFalseFalseFalseFalseFalseFalseFalseFalse
AFFX-TrpnX-5_atFalseFalseFalseFalseFalseFalseFalseFalse
AFFX-TrpnX-M_atFalseFalseFalseFalseFalseFalseFalseFalse
\n", "

45101 rows × 8 columns

\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 \\\n", "gene \n", "1415670_at True True True True \n", "1415671_at True True True True \n", "1415672_at True True True True \n", "1415673_at False False False False \n", "1415674_a_at True True True True \n", "... ... ... ... ... \n", "AFFX-TransRecMur/X57349_5_at False False False False \n", "AFFX-TransRecMur/X57349_M_at False False False False \n", "AFFX-TrpnX-3_at False False False False \n", "AFFX-TrpnX-5_at False False False False \n", "AFFX-TrpnX-M_at False False False False \n", "\n", " Heart_KO_1 Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at True True False True \n", "1415671_at True True True True \n", "1415672_at True True True True \n", "1415673_at False False False False \n", "1415674_a_at True True True True \n", "... ... ... ... ... \n", "AFFX-TransRecMur/X57349_5_at False False False False \n", "AFFX-TransRecMur/X57349_M_at False False False False \n", "AFFX-TrpnX-3_at False False False False \n", "AFFX-TrpnX-5_at False False False False \n", "AFFX-TrpnX-M_at False False False False \n", "\n", "[45101 rows x 8 columns]" ] }, "execution_count": 50, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df>1000" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Of course, if we try to make a data frame based on this filter we will have a dataframe with a lot of NaN (Not a Number) entries:" ] }, { "cell_type": "code", "execution_count": 51, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.4471182.4641206.2261196.0301174.6181184.458NaN1214.540
1415671_at3490.0982882.7842650.0332934.8612723.9762823.2602721.8842790.834
1415672_at4510.3694292.0574071.0574275.2764127.9414045.9904553.7364358.635
1415673_atNaNNaNNaNNaNNaNNaNNaNNaN
1415674_a_at1400.3251328.2951416.9231388.4181459.9561462.9641237.2441797.906
...........................
AFFX-TransRecMur/X57349_5_atNaNNaNNaNNaNNaNNaNNaNNaN
AFFX-TransRecMur/X57349_M_atNaNNaNNaNNaNNaNNaNNaNNaN
AFFX-TrpnX-3_atNaNNaNNaNNaNNaNNaNNaNNaN
AFFX-TrpnX-5_atNaNNaNNaNNaNNaNNaNNaNNaN
AFFX-TrpnX-M_atNaNNaNNaNNaNNaNNaNNaNNaN
\n", "

45101 rows × 8 columns

\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 \\\n", "gene \n", "1415670_at 1214.447 1182.464 1206.226 1196.030 \n", "1415671_at 3490.098 2882.784 2650.033 2934.861 \n", "1415672_at 4510.369 4292.057 4071.057 4275.276 \n", "1415673_at NaN NaN NaN NaN \n", "1415674_a_at 1400.325 1328.295 1416.923 1388.418 \n", "... ... ... ... ... \n", "AFFX-TransRecMur/X57349_5_at NaN NaN NaN NaN \n", "AFFX-TransRecMur/X57349_M_at NaN NaN NaN NaN \n", "AFFX-TrpnX-3_at NaN NaN NaN NaN \n", "AFFX-TrpnX-5_at NaN NaN NaN NaN \n", "AFFX-TrpnX-M_at NaN NaN NaN NaN \n", "\n", " Heart_KO_1 Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1174.618 1184.458 NaN 1214.540 \n", "1415671_at 2723.976 2823.260 2721.884 2790.834 \n", "1415672_at 4127.941 4045.990 4553.736 4358.635 \n", "1415673_at NaN NaN NaN NaN \n", "1415674_a_at 1459.956 1462.964 1237.244 1797.906 \n", "... ... ... ... ... \n", "AFFX-TransRecMur/X57349_5_at NaN NaN NaN NaN \n", "AFFX-TransRecMur/X57349_M_at NaN NaN NaN NaN \n", "AFFX-TrpnX-3_at NaN NaN NaN NaN \n", "AFFX-TrpnX-5_at NaN NaN NaN NaN \n", "AFFX-TrpnX-M_at NaN NaN NaN NaN \n", "\n", "[45101 rows x 8 columns]" ] }, "execution_count": 51, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[df>1000]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "A useful function to remove rows that contain NAN values is [dropna](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.dropna.html). We could use it to get a list of genes that have some minimal expression in all of the samples:" ] }, { "cell_type": "code", "execution_count": 52, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415671_at3490.0982882.7842650.0332934.8612723.9762823.2602721.8842790.834
1415672_at4510.3694292.0574071.0574275.2764127.9414045.9904553.7364358.635
1415674_a_at1400.3251328.2951416.9231388.4181459.9561462.9641237.2441797.906
1415676_a_at7927.8976705.2787242.7066859.5698368.4348531.3298547.2477839.285
1415678_at5476.5124205.5165146.7314337.7335979.4295316.2605421.7475278.323
...........................
AFFX-r2-Ec-bioC-5_at4600.5554277.7524360.8194568.1325267.0904405.1714139.6613921.692
AFFX-r2-Ec-bioD-3_at24437.63022834.85023117.29022103.41026510.62022345.28023486.88022942.520
AFFX-r2-Ec-bioD-5_at22073.23021445.68020391.22019246.60025175.55019780.71020764.07020664.880
AFFX-r2-P1-cre-3_at70849.58077527.95066128.73056309.32085272.30067346.14075702.37066842.450
AFFX-r2-P1-cre-5_at53018.43058323.05049811.45044437.16066640.52050457.06056151.34049132.230
\n", "

5596 rows × 8 columns

\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 \\\n", "gene \n", "1415671_at 3490.098 2882.784 2650.033 2934.861 \n", "1415672_at 4510.369 4292.057 4071.057 4275.276 \n", "1415674_a_at 1400.325 1328.295 1416.923 1388.418 \n", "1415676_a_at 7927.897 6705.278 7242.706 6859.569 \n", "1415678_at 5476.512 4205.516 5146.731 4337.733 \n", "... ... ... ... ... \n", "AFFX-r2-Ec-bioC-5_at 4600.555 4277.752 4360.819 4568.132 \n", "AFFX-r2-Ec-bioD-3_at 24437.630 22834.850 23117.290 22103.410 \n", "AFFX-r2-Ec-bioD-5_at 22073.230 21445.680 20391.220 19246.600 \n", "AFFX-r2-P1-cre-3_at 70849.580 77527.950 66128.730 56309.320 \n", "AFFX-r2-P1-cre-5_at 53018.430 58323.050 49811.450 44437.160 \n", "\n", " Heart_KO_1 Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415671_at 2723.976 2823.260 2721.884 2790.834 \n", "1415672_at 4127.941 4045.990 4553.736 4358.635 \n", "1415674_a_at 1459.956 1462.964 1237.244 1797.906 \n", "1415676_a_at 8368.434 8531.329 8547.247 7839.285 \n", "1415678_at 5979.429 5316.260 5421.747 5278.323 \n", "... ... ... ... ... \n", "AFFX-r2-Ec-bioC-5_at 5267.090 4405.171 4139.661 3921.692 \n", "AFFX-r2-Ec-bioD-3_at 26510.620 22345.280 23486.880 22942.520 \n", "AFFX-r2-Ec-bioD-5_at 25175.550 19780.710 20764.070 20664.880 \n", "AFFX-r2-P1-cre-3_at 85272.300 67346.140 75702.370 66842.450 \n", "AFFX-r2-P1-cre-5_at 66640.520 50457.060 56151.340 49132.230 \n", "\n", "[5596 rows x 8 columns]" ] }, "execution_count": 52, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[df>1000].dropna()" ] }, { "cell_type": "markdown", "metadata": { "collapsed": true }, "source": [ "## 4. Useful methods of dataframes" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "[shape](http://pandas.pydata.org/pandas-docs/version/0.17.0/generated/pandas.DataFrame.shape.html), which shows the dimensionality of the DataFrame. It is not a method but an attribute of the dataframe object." ] }, { "cell_type": "code", "execution_count": 55, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(45101, 8)" ] }, "execution_count": 55, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.shape" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can also get the number of rows with the ubiquitous `len()` function:" ] }, { "cell_type": "code", "execution_count": 56, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "33473" ] }, "execution_count": 56, "metadata": {}, "output_type": "execute_result" } ], "source": [ "len(df[df['Heart_WT_1']>20])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or with the [`count()`](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.count.html) function (which will be applied to all columns):" ] }, { "cell_type": "code", "execution_count": 57, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Heart_WT_1 33473\n", "Heart_WT_2 33473\n", "Heart_WT_3 33473\n", "Heart_WT_4 33473\n", "Heart_KO_1 33473\n", "Heart_KO_2 33473\n", "Heart_KO_3 33473\n", "Heart_KO_4 33473\n", "dtype: int64" ] }, "execution_count": 57, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df[df['Heart_WT_1']>20].count()" ] }, { "cell_type": "markdown", "metadata": { "collapsed": true }, "source": [ "### 4.1 Applying functions to dataframe rows or columns" ] }, { "cell_type": "markdown", "metadata": { "collapsed": true }, "source": [ "We can write our own functions and `apply` them row-wise or column-wise to our dataframe. \n", "\n", "For example, let's implement our version of the default [mean](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.mean.html) function.\n", "\n", "To see better what's going one, we'll extract a short section of our dataframe:" ] }, { "cell_type": "code", "execution_count": 152, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
1415675_at497.7774441.8907463.8198471.8565404.6885357.1436442.7815519.2141
1415676_a_at7927.89706705.27807242.70606859.56908368.43408531.32908547.24707839.2850
1415677_at781.6489513.3997481.5474646.0389695.0452768.4124574.2578521.8760
1415678_at5476.51204205.51605146.73104337.73305979.42905316.26005421.74705278.3230
1415679_at2931.39002882.16102844.51202858.87302905.79303177.12204169.55003241.5460
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "1415675_at 497.7774 441.8907 463.8198 471.8565 404.6885 \n", "1415676_a_at 7927.8970 6705.2780 7242.7060 6859.5690 8368.4340 \n", "1415677_at 781.6489 513.3997 481.5474 646.0389 695.0452 \n", "1415678_at 5476.5120 4205.5160 5146.7310 4337.7330 5979.4290 \n", "1415679_at 2931.3900 2882.1610 2844.5120 2858.8730 2905.7930 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 \n", "1415675_at 357.1436 442.7815 519.2141 \n", "1415676_a_at 8531.3290 8547.2470 7839.2850 \n", "1415677_at 768.4124 574.2578 521.8760 \n", "1415678_at 5316.2600 5421.7470 5278.3230 \n", "1415679_at 3177.1220 4169.5500 3241.5460 " ] }, "execution_count": 152, "metadata": {}, "output_type": "execute_result" } ], "source": [ "short_df = df.iloc[0:10,0:8]\n", "short_df" ] }, { "cell_type": "code", "execution_count": 153, "metadata": {}, "outputs": [], "source": [ "# we assume that we get an array/list/Series as input\n", "def calculate_mean(inp):\n", " \n", " '''Calculate mean'''\n", " \n", " total = 0\n", " for i in inp:\n", " total = total + i\n", " return total/len(inp)\n", " \n", " # or just\n", " # return np.average(inp)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now we apply the function across columns (axis = 1):" ] }, { "cell_type": "code", "execution_count": 154, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "gene\n", "1415670_at 1169.791663\n", "1415671_at 2877.216250\n", "1415672_at 4279.382625\n", "1415673_at 497.257538\n", "1415674_a_at 1436.503875\n", "1415675_at 449.896513\n", "1415676_a_at 7752.718125\n", "1415677_at 622.778288\n", "1415678_at 5145.281375\n", "1415679_at 3126.368375\n", "dtype: float64" ] }, "execution_count": 154, "metadata": {}, "output_type": "execute_result" } ], "source": [ "short_df.apply(calculate_mean, axis=1)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "which gives the same output as the default function:" ] }, { "cell_type": "code", "execution_count": 155, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "gene\n", "1415670_at 1169.791663\n", "1415671_at 2877.216250\n", "1415672_at 4279.382625\n", "1415673_at 497.257538\n", "1415674_a_at 1436.503875\n", "1415675_at 449.896513\n", "1415676_a_at 7752.718125\n", "1415677_at 622.778288\n", "1415678_at 5145.281375\n", "1415679_at 3126.368375\n", "dtype: float64" ] }, "execution_count": 155, "metadata": {}, "output_type": "execute_result" } ], "source": [ "short_df.mean(axis=1)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4.2 the groupby() function" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's say we want to group the genes by the direction of their change in the KO. But we'll take again a smaller part of the data frame for this:" ] }, { "cell_type": "code", "execution_count": 157, "metadata": {}, "outputs": [], "source": [ "short_df = df.iloc[0:10][['log2_Heart_WT_avg', 'log2_Heart_KO_avg', 'log2FC_KO_WT']]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now let's make an additional column that tells us the direction of the change: 'Upregulated' if the log2FC > 0, 'Downregulated' if the log2FC < 0 and 'Unchanged' if the log2FC == 0. Let's first define our own function for this:" ] }, { "cell_type": "code", "execution_count": 185, "metadata": {}, "outputs": [], "source": [ "def directionality(x):\n", " if(x > 0):\n", " return 'Upregulated'\n", " elif(x < 0):\n", " return 'Downregulated'\n", " else:\n", " return 'Unchanged'\n", " " ] }, { "cell_type": "code", "execution_count": 188, "metadata": {}, "outputs": [], "source": [ "short_df['Type']=short_df['log2FC_KO_WT'].map(directionality)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Then we can group by the type of change and count how many genes were of these three types:" ] }, { "cell_type": "code", "execution_count": 189, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
log2_Heart_WT_avglog2_Heart_KO_avglog2FC_KO_WTType2
Type
Downregulated4444
Upregulated6666
\n", "
" ], "text/plain": [ " log2_Heart_WT_avg log2_Heart_KO_avg log2FC_KO_WT Type2\n", "Type \n", "Downregulated 4 4 4 4\n", "Upregulated 6 6 6 6" ] }, "execution_count": 189, "metadata": {}, "output_type": "execute_result" } ], "source": [ "short_df.groupby('Type').count()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can do this on the large dataset as well (we print out the result on only one column):" ] }, { "cell_type": "code", "execution_count": 190, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/plain": [ "Type\n", "Downregulated 23967\n", "Upregulated 21134\n", "Name: Type, dtype: int64" ] }, "execution_count": 190, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df['Type']=df['log2FC_KO_WT'].map(directionality)\n", "df.groupby('Type')['Type'].count()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4.3 Write files to disk" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "At the end of the analysis we want to save our results so we should write some files out. \n", "\n", "We can use the function [to_csv](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.to_csv.html):" ] }, { "cell_type": "code", "execution_count": 191, "metadata": {}, "outputs": [], "source": [ "short_df.head().to_csv(\"short_example.csv\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or we can export it as tsv" ] }, { "cell_type": "code", "execution_count": 192, "metadata": {}, "outputs": [], "source": [ "short_df.head().to_csv(\"short_example.tsv\", sep='\\t')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or we can export it without the header" ] }, { "cell_type": "code", "execution_count": 193, "metadata": {}, "outputs": [], "source": [ "short_df.head().to_csv(\"short_example_no_header.tsv\", sep='\\t', header=None)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "One we have seen already, but we add an additional parameter" ] }, { "cell_type": "code", "execution_count": 53, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
1415675_at497.7774441.8907463.8198471.8565404.6885357.1436442.7815519.2141
1415676_a_at7927.89706705.27807242.70606859.56908368.43408531.32908547.24707839.2850
1415677_at781.6489513.3997481.5474646.0389695.0452768.4124574.2578521.8760
1415678_at5476.51204205.51605146.73104337.73305979.42905316.26005421.74705278.3230
1415679_at2931.39002882.16102844.51202858.87302905.79303177.12204169.55003241.5460
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "1415675_at 497.7774 441.8907 463.8198 471.8565 404.6885 \n", "1415676_a_at 7927.8970 6705.2780 7242.7060 6859.5690 8368.4340 \n", "1415677_at 781.6489 513.3997 481.5474 646.0389 695.0452 \n", "1415678_at 5476.5120 4205.5160 5146.7310 4337.7330 5979.4290 \n", "1415679_at 2931.3900 2882.1610 2844.5120 2858.8730 2905.7930 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 \n", "1415675_at 357.1436 442.7815 519.2141 \n", "1415676_a_at 8531.3290 8547.2470 7839.2850 \n", "1415677_at 768.4124 574.2578 521.8760 \n", "1415678_at 5316.2600 5421.7470 5278.3230 \n", "1415679_at 3177.1220 4169.5500 3241.5460 " ] }, "execution_count": 53, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head(n=10)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Exercises with pandas\n", "\n", "## Exercise 01\n", "\n", "Create a pandas dataframe where the index consists of the integers from 1-100. \n", "Remove a few rows from the dataframe. Slice the first rows using `.iloc` and `.loc` and make sure that you get the expected behavior.\n", "\n", "\n", "## Exercise 02\n", "\n", "Grab the most recent data on:\n", " * [number of hospitalised person](https://raw.githubusercontent.com/daenuprobst/covid19-cases-switzerland/master/covid19_hospitalized_switzerland_openzh.csv)\n", " * [canton demographics](https://raw.githubusercontent.com/daenuprobst/covid19-cases-switzerland/master/demographics.csv)\n", "\n", "1. download and read these two data files as pandas `DataFrame`\n", "2. Modify the numbers from \"number of people hospitalized\" to \"number of people hospitalized per 10 000 habitants\" for each canton\n", "3. get, for each canton, the date at which the hosptitalization rate was maximal\n", "\n", "## Exercise 03\n", "\n", "Work on the data from the `\"GSE41558_series_matrix.tsv\"` file\n", "1. Center each column : substract their mean from their values\n", "2. select from the data-frame only the genes whose expression is above the column-wise average in all the WT samples" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# If you want to dig deeper :" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 5. Sorting operations on dataframes\n", "\n", "Sort dataframe based on specific column(s):" ] }, { "cell_type": "code", "execution_count": 58, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1455997_a_at96713.59105301.3089181.6272801.18124739.0090422.16109169.8089733.48
1448826_at89797.6997492.8481668.7771079.91111231.5089459.20105711.3085422.31
1426088_at86414.3496914.3489977.7054913.4299651.9194240.30103890.6082828.70
1418726_a_at81302.7890228.4776994.1866917.99103520.6085074.2795528.8480444.54
1415927_at79998.3088502.8574298.6461849.0193866.8476351.3891672.9374601.72
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1455997_a_at 96713.59 105301.30 89181.62 72801.18 124739.00 \n", "1448826_at 89797.69 97492.84 81668.77 71079.91 111231.50 \n", "1426088_at 86414.34 96914.34 89977.70 54913.42 99651.91 \n", "1418726_a_at 81302.78 90228.47 76994.18 66917.99 103520.60 \n", "1415927_at 79998.30 88502.85 74298.64 61849.01 93866.84 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1455997_a_at 90422.16 109169.80 89733.48 \n", "1448826_at 89459.20 105711.30 85422.31 \n", "1426088_at 94240.30 103890.60 82828.70 \n", "1418726_a_at 85074.27 95528.84 80444.54 \n", "1415927_at 76351.38 91672.93 74601.72 " ] }, "execution_count": 58, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.sort_values('Heart_WT_1', ascending=False).head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or we can sort by index :" ] }, { "cell_type": "code", "execution_count": 59, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 59, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.sort_index(ascending=True).head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finding the minimum value in each column:" ] }, { "cell_type": "code", "execution_count": 60, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Heart_WT_1 0.520472\n", "Heart_WT_2 0.438004\n", "Heart_WT_3 0.440301\n", "Heart_WT_4 0.358835\n", "Heart_KO_1 0.218736\n", "Heart_KO_2 0.430182\n", "Heart_KO_3 0.286753\n", "Heart_KO_4 0.366639\n", "dtype: float64" ] }, "execution_count": 60, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.min(axis=0)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Note the `axis` parameter. This gives the dimension along which values are compared. `axis=0` indicates that the comparison is across rows and there looping over all index values in all other dimensions. In this case, for each column we got a value.\n", "\n", "Let's now find the min in each row (in this case for each gene), over all columns (samples):" ] }, { "cell_type": "code", "execution_count": 62, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/plain": [ "gene\n", "1415670_at 985.5503\n", "1415671_at 2650.0330\n", "1415672_at 4045.9900\n", "1415673_at 323.8668\n", "1415674_a_at 1237.2440\n", "dtype: float64" ] }, "execution_count": 62, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.min(axis=1).head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Maybe what we want is not the minimum value but the index at which it is found: " ] }, { "cell_type": "code", "execution_count": 65, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "gene\n", "1415670_at Heart_KO_3\n", "1415671_at Heart_WT_3\n", "1415672_at Heart_KO_2\n", "1415673_at Heart_KO_4\n", "1415674_a_at Heart_KO_3\n", "dtype: object" ] }, "execution_count": 65, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.idxmin(axis=1).head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or maybe we want both:" ] }, { "cell_type": "code", "execution_count": 71, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Heart_WT_1 (1439379_x_at, 0.5204723)\n", "Heart_WT_2 (1447645_x_at, 0.43800390000000006)\n", "Heart_WT_3 (1443834_at, 0.4403009)\n", "Heart_WT_4 (1429855_at, 0.3588345)\n", "Heart_KO_1 (1438333_at, 0.2187357)\n", "Heart_KO_2 (1430240_a_at, 0.4301823)\n", "Heart_KO_3 (1420239_x_at, 0.28675320000000004)\n", "Heart_KO_4 (1438524_x_at, 0.3666389)\n", "dtype: object" ] }, "execution_count": 71, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def myfilter(df):\n", " return df.idxmin(axis=0),df.min(axis=0)\n", " \n", "df.apply( myfilter )" ] }, { "cell_type": "code", "execution_count": 74, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Heart_WT_1 (1415670_at, 1214.447)\n", "Heart_WT_2 (1415670_at, 1182.464)\n", "Heart_WT_3 (1415670_at, 1206.226)\n", "Heart_WT_4 (1415670_at, 1196.03)\n", "Heart_KO_1 (1415670_at, 1174.618)\n", "Heart_KO_2 (1415670_at, 1184.458)\n", "Heart_KO_3 (1415670_at, 985.5503)\n", "Heart_KO_4 (1415670_at, 1214.54)\n", "dtype: object" ] }, "execution_count": 74, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# finding the gene with mininum expression, but on the first 3 rows only\n", "df[0:3].apply( myfilter )" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Note the application of a function that generates tuples of (index (gene name) - value). If the minimum occurs multiple times, only the first index where it occurs is returned." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 6. Extending a dataframe by adding new columns" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can set up a new dataframe and concatenate it to the original dataframe using the `concat` method:" ] }, { "cell_type": "code", "execution_count": 84, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_avgHeart_KO_avg
gene
1415670_at1199.7917501139.791575
1415671_at2989.4440002764.988500
1415672_at4287.1897504271.575500
1415673_at489.157575505.357500
1415674_a_at1383.4902501489.517500
\n", "
" ], "text/plain": [ " Heart_WT_avg Heart_KO_avg\n", "gene \n", "1415670_at 1199.791750 1139.791575\n", "1415671_at 2989.444000 2764.988500\n", "1415672_at 4287.189750 4271.575500\n", "1415673_at 489.157575 505.357500\n", "1415674_a_at 1383.490250 1489.517500" ] }, "execution_count": 84, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dfavg = pd.DataFrame()\n", "dfavg['Heart_WT_avg'] = (df['Heart_WT_1']+df['Heart_WT_2']+df['Heart_WT_3']+df['Heart_WT_4'])/4\n", "dfavg['Heart_KO_avg'] = (df['Heart_KO_1']+df['Heart_KO_2']+df['Heart_KO_3']+df['Heart_KO_4'])/4\n", "dfavg.head()" ] }, { "cell_type": "code", "execution_count": 85, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2AvgHeart_WT_avgHeart_KO_avg
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.1920361199.7917501139.791575
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.4904582989.4440002764.988500
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.0631874287.1897504271.575500
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.957849489.157575505.357500
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.4883461383.4902501489.517500
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \\\n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 \n", "\n", " Heart_WT_avg Heart_KO_avg \n", "gene \n", "1415670_at 1199.791750 1139.791575 \n", "1415671_at 2989.444000 2764.988500 \n", "1415672_at 4287.189750 4271.575500 \n", "1415673_at 489.157575 505.357500 \n", "1415674_a_at 1383.490250 1489.517500 " ] }, "execution_count": 85, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dfavg = pd.DataFrame()\n", "dfavg['Heart_WT_avg'] = (df['Heart_WT_1']+df['Heart_WT_2']+df['Heart_WT_3']+df['Heart_WT_4'])/4\n", "dfavg['Heart_KO_avg'] = (df['Heart_KO_1']+df['Heart_KO_2']+df['Heart_KO_3']+df['Heart_KO_4'])/4\n", "\n", "dfall = pd.concat([df, dfavg], axis=1)\n", "dfall.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or we can modify the original dataframe in place, adding columns to it:" ] }, { "cell_type": "code", "execution_count": 86, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2Avg
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.192036
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.490458
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.063187
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.957849
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.488346
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 " ] }, "execution_count": 86, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df['Heart_WT_avg'] = (df['Heart_WT_1']+df['Heart_WT_2']+df['Heart_WT_3']+df['Heart_WT_4'])/4\n", "df['Heart_KO_avg'] = (df['Heart_KO_1']+df['Heart_KO_2']+df['Heart_KO_3']+df['Heart_KO_4'])/4\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 7. Use of numpy functions with pandas dataframes" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's say we want to calculate the log average expression value. We could do it like this:" ] }, { "cell_type": "code", "execution_count": 87, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2Avg
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.192036
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.490458
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.063187
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.957849
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.488346
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 " ] }, "execution_count": 87, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df['log2_Heart_KO_avg'] = np.log2(df['Heart_KO_avg'])\n", "df['log2_Heart_WT_avg'] = np.log2(df['Heart_WT_avg'])\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "And now we can calculate the log fold-change in expression between conditions:" ] }, { "cell_type": "code", "execution_count": 88, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2Avglog2FC
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.192036-0.074014
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.490458-0.112604
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.063187-0.005264
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.9578490.047005
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.4883460.106533
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \\\n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 \n", "\n", " log2FC \n", "gene \n", "1415670_at -0.074014 \n", "1415671_at -0.112604 \n", "1415672_at -0.005264 \n", "1415673_at 0.047005 \n", "1415674_a_at 0.106533 " ] }, "execution_count": 88, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df['log2FC'] = df['log2_Heart_KO_avg']-df['log2_Heart_WT_avg']\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "But now we realized that we didn't keep track of how we calculated the fold change (which was the numerator/denominator). We can fix this:" ] }, { "cell_type": "code", "execution_count": 89, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2Avglog2FC
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.192036-0.074014
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.490458-0.112604
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.063187-0.005264
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.9578490.047005
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.4883460.106533
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \\\n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 \n", "\n", " log2FC \n", "gene \n", "1415670_at -0.074014 \n", "1415671_at -0.112604 \n", "1415672_at -0.005264 \n", "1415673_at 0.047005 \n", "1415674_a_at 0.106533 " ] }, "execution_count": 89, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df['log2FC_KO_WT'] = df['log2_Heart_KO_avg']-df['log2_Heart_WT_avg']\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "And then we have to remove the old column, with the `.drop()` method:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "df.drop('log2FC', axis=1, inplace=True)\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally, let's add one more column with the average expression across all samples" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "df['log2Avg'] = np.log2((df['Heart_KO_avg']+df['Heart_WT_avg'])/2)\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 9. Plotting with pandas and matplotlib" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "Now let's explore our data a bit. First, a matrix of scatter plots for all pairwise sample comparisons:" ] }, { "cell_type": "code", "execution_count": 93, "metadata": {}, "outputs": [], "source": [ "import matplotlib.pyplot as plt" ] }, { "cell_type": "code", "execution_count": 94, "metadata": { "scrolled": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "from pandas.plotting import scatter_matrix\n", "\n", "scatter_matrix(np.log2(df.iloc[:,0:8]), figsize=(12, 12), diagonal='kde')\n", "plt.show() # this line makes the plot appear" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The so-called M-A plot of expression change between KO and WT relative to average expression in these two conditions:" ] }, { "cell_type": "code", "execution_count": 95, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "pd.DataFrame(df[['log2FC_KO_WT', 'log2Avg']]).plot(x='log2Avg', y='log2FC_KO_WT', kind='scatter')\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Histogram of the log fold changes" ] }, { "cell_type": "code", "execution_count": 96, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "df['log2FC_KO_WT'].hist(bins=100)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Or we can check how well the expression changes fit a gaussian distribution" ] }, { "cell_type": "code", "execution_count": 103, "metadata": { "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import scipy.stats as stats\n", "\n", "# best fit of data with a normal law\n", "(mu, sigma) = stats.norm.fit(df['log2FC_KO_WT'])\n", "\n", "# the histogram of the data\n", "n, bins, patches = plt.hist(df['log2FC_KO_WT'], 60, density=1, facecolor='blue', alpha=1)\n", "\n", "y = stats.norm.pdf( bins, loc = mu , scale = sigma )\n", "# add a 'best fit' line\n", "l = plt.plot(bins, y, 'r--', linewidth=2)\n", "\n", "#plot\n", "plt.xlabel('Log2 fold change')\n", "plt.ylabel('Probability')\n", "plt.title(r'$\\mathrm{Histogram\\ of\\ log2 fold change:}\\ \\mu=%.3f,\\ \\sigma=%.3f$' %(mu, sigma))\n", "plt.grid(True)\n", "\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 10. Recap of the methods we have covered so far" ] }, { "cell_type": "code", "execution_count": 107, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2Avg
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.192036
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.490458
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.063187
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.957849
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.488346
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 " ] }, "execution_count": 107, "metadata": {}, "output_type": "execute_result" } ], "source": [ "column_names = [\"gene\", \"Heart_WT_1\", \"Heart_WT_2\", \"Heart_WT_3\", \"Heart_WT_4\", \"Heart_KO_1\", \"Heart_KO_2\", \"Heart_KO_3\", \"Heart_KO_4\"]\n", "df = pd.read_csv(\"GSE41558_series_matrix_no_header.tsv\", sep='\\t', header=None)\n", "df.columns = column_names\n", "df.set_index('gene', inplace=True) \n", "df.head()\n", "df['Heart_WT_avg'] = (df['Heart_WT_1']+df['Heart_WT_2']+df['Heart_WT_3']+df['Heart_WT_4'])/4\n", "df['Heart_KO_avg'] = (df['Heart_KO_1']+df['Heart_KO_2']+df['Heart_KO_3']+df['Heart_KO_4'])/4\n", "df['log2_Heart_KO_avg'] = np.log2(df['Heart_KO_avg'])\n", "df['log2_Heart_WT_avg'] = np.log2(df['Heart_WT_avg'])\n", "df['log2FC_KO_WT'] = np.log2(df['Heart_KO_avg'])-np.log2(df['Heart_WT_avg'])\n", "df['log2Avg'] = np.log2((df['Heart_KO_avg']+df['Heart_WT_avg'])/2)\n", "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 11. Merge and join" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "[Merge](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.merge.html) or [join](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.join.html) operations combine data sets, linking rows based on their keys. \n", "\n", "Here's how we construct a dataframe from a dictionary data structure, where dictionary keys are treated as column names, list of values associated with a key is treated as list of elements in the corresponding column, and rows are contructed based on the index of elements within the list of elements in the column (note however that all columns should have the same length):" ] }, { "cell_type": "code", "execution_count": 108, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
keydata1
0b0
1b1
2a2
3c3
4a4
5a5
6b6
\n", "
" ], "text/plain": [ " key data1\n", "0 b 0\n", "1 b 1\n", "2 a 2\n", "3 c 3\n", "4 a 4\n", "5 a 5\n", "6 b 6" ] }, "execution_count": 108, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df1 = pd.DataFrame({'key': ['b','b','a','c','a','a','b'], 'data1': range(7)})\n", "#print(type(df1))\n", "df1" ] }, { "cell_type": "code", "execution_count": 109, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
keydata2
0a0
1b1
2d2
\n", "
" ], "text/plain": [ " key data2\n", "0 a 0\n", "1 b 1\n", "2 d 2" ] }, "execution_count": 109, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df2 = pd.DataFrame({'key': ['a','b','d'], 'data2': range(3)})\n", "df2" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's merge the two data frames, with the default application of the `merge` function:" ] }, { "cell_type": "code", "execution_count": 110, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
keydata1data2
0b01
1b11
2b61
3a20
4a40
5a50
\n", "
" ], "text/plain": [ " key data1 data2\n", "0 b 0 1\n", "1 b 1 1\n", "2 b 6 1\n", "3 a 2 0\n", "4 a 4 0\n", "5 a 5 0" ] }, "execution_count": 110, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pd.merge(df1, df2)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "How has python interpreted our call to `merge`?\n", "\n", "1. It has assumed that we want to merge on the basis of the common `key` column.\n", "\n", "2. It has identified the values of `key` which occur in both dataframes\n", "\n", "3. It has generated a dataframe with all combinations of rows from dataframes 1 and 2 that are associated with a particular key value" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can be more precise ourselves in specifying how to merge the dataframes, using the **on** option:" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [ "pd.merge(df1, df2, on='key')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "By default, merge performs an 'inner' operation, taking the intersection of the key sets. However, we can specify the way we want to merge through options like 'outer', 'left', 'right'. This determines which set of keys to consider (the union of the two sets, all of those that occur in df1, all of those that occur in df2). Missing values show up as NaN." ] }, { "cell_type": "code", "execution_count": 111, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
keydata1data2
0b0.01.0
1b1.01.0
2b6.01.0
3a2.00.0
4a4.00.0
5a5.00.0
6c3.0NaN
7dNaN2.0
\n", "
" ], "text/plain": [ " key data1 data2\n", "0 b 0.0 1.0\n", "1 b 1.0 1.0\n", "2 b 6.0 1.0\n", "3 a 2.0 0.0\n", "4 a 4.0 0.0\n", "5 a 5.0 0.0\n", "6 c 3.0 NaN\n", "7 d NaN 2.0" ] }, "execution_count": 111, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pd.merge(df1, df2, on='key', how='outer')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Merging can also be done based on index. For this lets use our example that we had before. " ] }, { "cell_type": "code", "execution_count": 113, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4Heart_WT_avgHeart_KO_avglog2_Heart_KO_avglog2_Heart_WT_avglog2FC_KO_WTlog2Avg
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.54001199.7917501139.79157510.15455410.228568-0.07401410.192036
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.83402989.4440002764.98850011.43305811.545661-0.11260411.490458
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.63504287.1897504271.57550012.06055312.065817-0.00526412.063187
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668489.157575505.3575008.9811618.9341550.0470058.957849
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.90601383.4902501489.51750010.54062910.4340970.10653310.488346
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 Heart_WT_avg Heart_KO_avg \\\n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 1199.791750 1139.791575 \n", "1415671_at 2823.2600 2721.8840 2790.8340 2989.444000 2764.988500 \n", "1415672_at 4045.9900 4553.7360 4358.6350 4287.189750 4271.575500 \n", "1415673_at 544.3807 569.1154 323.8668 489.157575 505.357500 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 1383.490250 1489.517500 \n", "\n", " log2_Heart_KO_avg log2_Heart_WT_avg log2FC_KO_WT log2Avg \n", "gene \n", "1415670_at 10.154554 10.228568 -0.074014 10.192036 \n", "1415671_at 11.433058 11.545661 -0.112604 11.490458 \n", "1415672_at 12.060553 12.065817 -0.005264 12.063187 \n", "1415673_at 8.981161 8.934155 0.047005 8.957849 \n", "1415674_a_at 10.540629 10.434097 0.106533 10.488346 " ] }, "execution_count": 113, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's create two data frames, one containing the data for the WT and the other for the KO:" ] }, { "cell_type": "code", "execution_count": 116, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300\n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610\n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760\n", "1415673_at 598.8334 385.0178 458.4872 514.2919\n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180\n", " Heart_KO_1 Heart_KO_2 Heart_KO_3 Heart_KO_4\n", "gene \n", "1415670_at 1174.6180 1184.4580 985.5503 1214.5400\n", "1415671_at 2723.9760 2823.2600 2721.8840 2790.8340\n", "1415672_at 4127.9410 4045.9900 4553.7360 4358.6350\n", "1415673_at 584.0671 544.3807 569.1154 323.8668\n", "1415674_a_at 1459.9560 1462.9640 1237.2440 1797.9060\n" ] } ], "source": [ "wt_cols = [col for col in df.columns if 'WT' in col and 'avg' not in col and 'log' not in col]\n", "ko_cols = [col for col in df.columns if 'KO' in col and 'avg' not in col and 'log' not in col]\n", "df_WT = df[wt_cols]\n", "df_KO = df[ko_cols]\n", "print(df_WT.head())\n", "print(df_KO.head())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's merge these frames based on the index:" ] }, { "cell_type": "code", "execution_count": 117, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Heart_WT_1Heart_WT_2Heart_WT_3Heart_WT_4Heart_KO_1Heart_KO_2Heart_KO_3Heart_KO_4
gene
1415670_at1214.44701182.46401206.22601196.03001174.61801184.4580985.55031214.5400
1415671_at3490.09802882.78402650.03302934.86102723.97602823.26002721.88402790.8340
1415672_at4510.36904292.05704071.05704275.27604127.94104045.99004553.73604358.6350
1415673_at598.8334385.0178458.4872514.2919584.0671544.3807569.1154323.8668
1415674_a_at1400.32501328.29501416.92301388.41801459.95601462.96401237.24401797.9060
\n", "
" ], "text/plain": [ " Heart_WT_1 Heart_WT_2 Heart_WT_3 Heart_WT_4 Heart_KO_1 \\\n", "gene \n", "1415670_at 1214.4470 1182.4640 1206.2260 1196.0300 1174.6180 \n", "1415671_at 3490.0980 2882.7840 2650.0330 2934.8610 2723.9760 \n", "1415672_at 4510.3690 4292.0570 4071.0570 4275.2760 4127.9410 \n", "1415673_at 598.8334 385.0178 458.4872 514.2919 584.0671 \n", "1415674_a_at 1400.3250 1328.2950 1416.9230 1388.4180 1459.9560 \n", "\n", " Heart_KO_2 Heart_KO_3 Heart_KO_4 \n", "gene \n", "1415670_at 1184.4580 985.5503 1214.5400 \n", "1415671_at 2823.2600 2721.8840 2790.8340 \n", "1415672_at 4045.9900 4553.7360 4358.6350 \n", "1415673_at 544.3807 569.1154 323.8668 \n", "1415674_a_at 1462.9640 1237.2440 1797.9060 " ] }, "execution_count": 117, "metadata": {}, "output_type": "execute_result" } ], "source": [ "merged_df = pd.merge(df_WT, df_KO, left_index=True, right_index=True)\n", "merged_df.head()" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.4" } }, "nbformat": 4, "nbformat_minor": 2 }