Fig. 1

Conceptual overview of GOPredict. In-house and curated data (left) are used to create a gene-by-study matrix of ranks which is stored in the knowledge-base (large blue box). GOPredict uses the genewise study ranks to calculate gene K-ranks (yellow box, left). K-ranks are used to calculate cancer-essentiality for GO processes (yellow box, middle). K-ranks are recalibrated with GO process scores and then used to prioritize drugs and stratify samples for input query data sets