PredictSNP 2.0

Thistool estimates the deleteriousness of single-nucleotide mutations inthe context of the development of Mendelian diseases. The predictionsare based on the results of existing tools: CADD, DANN, FATHMM,FunSeq2 and GWAVA. To achieve the highest possible accuracy,developed consensual functions based on category optimal decisionthresholds differ according to the category of variants. Thesegeneral categories are recognized: (i) regulatory, (ii) splicing,(iii) synonymous, (iv) missense and (v)nonsense variants. Theevaluation on large datasets revealed a marked benefit of thisapproach while the web interface provides easily interpretableresults for all individual tools and our consensual predictiontogether with the links to the relevant databases and on-lineservices.