Fig. 1
From: Convert your favorite protein modeling program into a mutation predictor: “MODICT”
Classification of S m p d1G506R. a. Wildtype (blue), Smpd1G506R (red) and Smpd1V36A (orange) models are shown. The original position of glycine in wildtype, the substitution site in Smpd1G506R and the alanine 36 in Smpd1V36A are marked with gray arrows. Models have been further refined using the MODREFINER. A negative control score was generated by superimposing the refined wildtype on the initial wildtype whereas a known benign score was generated by superimposing the refined Smpd1V36A on the initial wildtype. A score for the test mutation was generated in the same manner. MODICT scores were generated taking into account the entire backbone (residues 1-629). b. Thresholds were calculated as shown in the right and the G506R mutation was classified based on the calculated score bracket as shown in the left. The value of kappa can be updated using the roc.pl script. (σ=standard deviation of S I and S C )