Human epidermal growth factor receptor (EGFR) has become a well-established target for the treatment of patients with non-small cell lung cancer (NSCLC). However, a large number of somatic mutations in such protein have been observed to cause drug resistance or sensitivity during pathological progression, limiting the application of reversible EGFR tyrosine kinase inhibitor therapy in NSCLC. In the current work, we describe an integration of in silico analysis and in vitro assay to profile six representative EGFR inhibitors against a panel of 71 observed somatic mutations in EGFR tyrosine kinase domain. In the procedure, the changes in interaction free energy of inhibitors with EGFR upon various mutations were calculated one by one using a rigorous computational scheme, which was preoptimized based on a set of structure-solved, affinity-known samples to improve its performance in characterizing the EGFR-inhibitor system. This method was later demonstrated to be effective in inferring drug response to the classical L858R and G719S mutations that confer constitutive activation for the EGFR kinase. It is found that the Staurosporine, a natural product isolated from the bacterium Streptomyces staurosporeus, exhibits selective inhibitory activity on the T790M and T790M/L858R mutants. This finding was subsequently solidified by in vitro kinase assay experiment; the inhibitory IC50 values of Staurosporine against wild-type, T790M and T790M/L858R mutant EGFR were measured to be 937, 12 and 3 nM, respectively.