Epidermal Growth Factor Receptor (EGFR) is a receptor tyrosine kinase protein present on the cell surface, which is mainly expressed in epithelial cells. It gets activated upon binding to its specific ligands like EGF (Epidermal Growth Factor), TGF (Transforming Growth Factor) and induces down-stream signalling pathways that ultimately lead to cell differentiation and proliferation.
In the current study, we mainly focused on EGFR point mutations which are more frequently observed in different cancer types. We classified the Single Amino acid Polymorphisms (SAPs) into disease-causing (driver) and neutral (passenger) mutations by machine learning approaches. The proposed method achieved classification accuracy of 80.2%, 81.8%, 77.9% and 75.14% for helix, strand, coil buried and exposed mutants, respectively.