CancerGD provides a searchable collection of genetic vulnerabilities associated with the alteration of driver genes in cancer cell lines. We integrate genotypic data describing the mutation or copy number alteration of driver genes with large scale loss of function screens in cancer cell lines. This allows us to identify instances where alteration of a specific driver gene is associated with an increased sensitivity to the perturbation of another gene. Currently the loss-of-function screens we integrate are generated using RNA interference (RNAi) based approaches but we anticipate the integration of CRISPR screens as they become available.
Our goal with this resource is to make the identification and interpretation of genotype-associated vulnerabilities from loss-of-function screens as easy as possible. To this end we integrate two additional data sources with the identified genetic dependencies. To identify potentially 'druggable' vulnerabilities we annotate each gene identified as a dependency according to whether it has an available inhibitor. This data is taken from DGIdb. To aid the interpretation of genetic dependencies we integrate known functional interactions from the STRING database. This allows the user to easily identify genetic dependencies where there is a known relationship between the dependency and the driver gene (e.g. ERBB2 amplification is associated with an increased dependency on the ERBB2 binding partner ERBB3). It is also possible to view the functional interactions between the dependencies associated with a given driver gene, to see if they are functionally related (e.g. SMAD4 loss is associated with an increased dependence upon a densely connected group of genes all involved in the mitotic checkpoint).
A preprint describing CancerGD.org is available here.
This project was supported by funding from The Health Research Board, Ireland.