Next generation Sequencing technologies deliver huge amounts of short sequences for a relative low price. The data is frequently of low quality (noise and artifacts) and must be cleaned before usage; the assembly steps are often difficult due to the nature of the sequences (repeats, GC content), the quality of the tools and can be very different with a reference (by mapping) or without a reference (de novo). The final results must be evaluated with various criteria. The aim of the course are: (i) learn how to measure the quality of the data, clean and extract the good information. (ii) learn to assemble a genome by mapping, based on a known reference (e.g., MAQ, Bowtie, BWA). (iii) learn to assemble a genome de novo (without a reference) (e.g., SOAPdenovo, ABySS, Velvet).