MAP Program

Through comprehensive analysis of population-level genetic sequencing data, we identify oncogenic mutations among hundreds of unique alterations within a single gene. Our MAP platform algorithm uses genetic and proteomic features to rank mutations for potential oncogenicity. We use our algorithm as a machine learning tool to classify mutations as either pathogenic or benign and predict the probability, or MAP score, that a mutation is pathogenic.

We confirm the oncogenicity of the identified mutations through cell and tumor models and reveal how these mutations drive conformational changes in proteins. This allows us to group subsets of mutations into families based upon similar protein structures and shared selectivity profiles.

Using these shared characteristics, we seek to develop single small molecule product candidates, each designed to inhibit only the intended mutation family.