Cost-effective mapping of genetic interactions in mammalian cells

Comprehensive maps of genetic interactions in mammalian cells are daunting to construct because of the large number of potential interactions, ~ 2 x 108 for protein coding genes. We previously used co-inheritance of distant genes from published radiation hybrid (RH) datasets to identify genetic interactions. However, it was necessary to combine six legacy datasets from four species to obtain adequate statistical power. Mapping resolution was also limited by the low density PCR genotyping. Here, we employ shallow sequencing of nascent human RH clones as an economical approach to constructing interaction maps. In this initial study, 15 clones were analyzed, enabling construction of a network with 225 genes and 2359 interactions (FDR < 0.05). Despite its small size, the network showed significant overlap with the previous RH network and with a protein-protein interaction network. Consumables were [lsim] $50 per clone, showing that affordable, high quality genetic interaction maps are feasible in mammalian cells.