Fig. 7

Suppressor gene prediction and validation. A A suppressor gene prediction model was developed using a random forest classifier. For each query gene, the rank of the validated suppressor gene(s) was determined on both a random gene list and on a list of genes ranked by the likeliness of being a suppressor gene using the prediction algorithm. The rank of the validated suppressor gene was plotted against the number of query genes that interacted with a suppressor gene with that rank. B FANCA knockout cells are sensitive to cisplatin. The indicated cell lines were treated with cisplatin for six days, after which cells were counted. Shading represents the standard error of the mean of at least three independent biological replicates. C Experimentally identified suppressor genes of FANCA. Boxplot showing the normalized read count for guide RNAs targeting the indicated suppressor genes after 18 days of incubation with a concentration of cisplatin that inhibits proliferation by ~ 80%. Knockout of the indicated genes specifically suppresses the proliferation defect of FANCA knockout cells. D Comparison of the median rank of confirmed suppressor genes, either in a list of genes ranked by the likeliness of being a suppressor gene using the random forest classifier or in a random gene list. Statistical significance was determined using Mann–Whitney U tests. * p < 0.05, ** p < 0.005, *** p < 0.0005. Horizontal lines in boxplots: median