Résumé
First, biologists, often without extensive computer training background, have to treat the more or less dirty raw sequence outputs from sequencers (fastq files from the loaded library) to get clean individual genomes or individual genotypes that can be further analysed with population genetic tools. This may consist in a few straightforward analyses or to much more complexe pipelines with potentially many input options to play with, depending on the sequencing technology (e.g. RADseq vs shotgun sequencing). The first part of my talk aims at describing the state of the art of this "sequence treatment in Ecology and Evolution" with a special emphasis on the need to develop flexible standard workflows with many check outputs to ensure reliable generation of population-genetics datasets by non-specialists in bio-informatics. Then, once the clean data set has been generated, most population genetics analyses requires complex softwares implementing cutting-edge statistical tools. Again, those software will be used by non-statisticians non-informaticians and thus requires to be developed in a (FAIR) robust manner. This will not be the second part of my talk...