Clustering: improve table_clstr.txt

Pas prioritaire. Make 1 line by cluster with list of associated genes separated by comas (and not 1 line = 1 cluster and 1 associated gene).