Skip to main content

Table 2 Methods for predicting the consequences of point mutations

From: Getting personalized cancer genome analysis into the clinic: the challenges in bioinformatics

Name URL How it works
SIFT Uses sequence homology scores that are calculated using position-specific scoring matrices with Dirichlet priors
Polyphen 2 Uses sequence conservation, structure and Swiss-Prot annotations
PMUT Formulates predictions with neural networks, using internal databases, secondary structure prediction and sequence conservation
SNPs3D Based on a support vector machine that uses structural or sequence conservation parameters
PantherPSEC19 Uses sequence homology scores calculated using PANTHER hidden Markov model families
Mutationassessor Provides predictions using additional information based on the specific patterns of conservation of protein families
VEP (Variant Effect Predictor) This system categorizes Ensembl genomic variants in known transcripts by their potential effect
KinMut Prediction of the consequences of mutations in protein kinases; the system was trained with specific information about the kinase subfamilies, and together with the predictions provides general information about the corresponding proteins, a comparison with other predictors and links to the related literature