Table 1.

FUSIL categories.

Mouse category	Human cell line category	Number of genes	% Overlap	FUSIL category
Lethal	Essential	413	35.09%	Cellular lethal (CL)
Lethal	Non-essential	764	64.91%	Developmental lethal (DL)
Subviable	Essential	16	3.66%	—
Subviable	Non-essential	421	96.34%	Subviable (SV)
Viable with phenotypic abnormalities	Essential	18	0.95%	—
Viable with phenotypic abnormalities	Non-essential	1867	99.05%	Viable with phenotype (VP)
Viable with normal phenotype	Essential	2	0.62%	—
Viable with normal phenotype	Non-essential	318	99.38%	Viable with no phenotype (VN)

Integration of data from human cell essentiality screens from the Avana data set and mouse phenotypes from IMPC screens for 4446 protein-coding genes that have data in both resources and a high-quality orthologue. This defined five mutually exclusive categories of intolerance to loss of function and the number of human protein-coding genes is shown for each. For 627 of the viable mouse lines, the number of procedures with QCed data available was <50% and thus they were classified as Viable with insufficient procedures (see “Methods”, Supplementary Table 1) and not incorporated into these FUSIL categories. The Viable with phenotype (VP) category indicates that the phenotypes of the knockout (loss of function) mouse line differ significantly from the wild-type mice in at least one of the many parameters measured as part of the IMPC phenotyping pipeline (average of 163 parameters measured on any given mouse).