Statistical transcriptomics data analysis: from design to reproducible research
Fribourg, May 1st, 8, 22, 29, 2015 from 10:15 to 12
OverviewThis "Chapitres Choisis" is part of a series of topics chosen for Masters students in life sciences of the University of Fribourg.May 1: Experimental design (FS)May 8: Normalization, batch effects, how to compare values (CS)May 22: "High dimensionality", why it may be a problem and what one can do or has to do. (CS)May 29: Reproducible research + exam (FS)
The goal of this course is to review good practices and classical mistakes when preparing and analyzing data of a transcriptomics experiment (e.g., RNAseq).
Often researchers proposing a transcriptomics experiment fail to think in advance about potential caveats that will render their data useless to answer their question. In this course, the participants will learn how to set up an experiment correctly and how to anticipate potential issues in the data analysis.
Working with large numbers of samples or data points can sometimes become too complex for classical statistics normally applied to small numbers. Here the participants will learn the limits and the advantages of different statistical methods, as well as strategies to correct for these issues.
Finally producing data and analyzing it with reproducibility in mind is also a critical point. What kind of data should you publish? How to describe the methodology of analysis? Versioning of software? Many other questions will be answered to help the participants make their experiment reproducible.
This course is open to Masters students and PhD students.
It is recommended to have basic statistics knowledge (e.g., t-test)
Université de Fribourg
Building Pérolles PER07, room 1.311
For technical and scientific questions, please contact email@example.com