The SIFT web server predicts the effects of amino acid substitutions on protein function. First introduced in 2001, SIFT has become a standard tool for characterizing missense variation. The authors have updated the SIFT genome-wide prediction tool since their 2009 publication and added new features to the insertion/deletion (indel) tool. They also show accuracy metrics on independent data sets. The original developers host the SIFT web server at FHCRC, JCVI, and it is currently located at BII. The URL is http://sift-dna.org. SIFT uses sequence homology to compute the likelihood that an amino acid substitution will have an adverse effect on protein function. The algorithm was extended to predict the effects of frameshifting indels. SIFT has been used in human genetic research, including cancer, Mendelian diseases, and infectious diseases. It has also been used to study the effects of missense mutations in agricultural plants and model organisms. The SIFT algorithm was tested on the HumVar and HumDiv data sets, which were created by the authors of PolyPhen-2. The performance of SIFT was assessed using various protein databases, with UniRef90 chosen for its high coverage, sensitivity, and balanced performance. SIFT exhibited higher specificity on the HumDiv and HumVar data sets. New features of the web server include a genome-wide database of predictions for nonsynonymous variants and variant annotation with dbSNP and 1000 Genomes. The server also supports file format conversion, including VCF, Pileup, and GFF files. The indel prediction tool was developed to predict the effects of frameshifting indels, achieving 84% accuracy, 90% sensitivity, and 81% precision. The SIFT web server is freely available to the academic community. The source code and Linux executables are publicly available, and a public image is available on the Amazon cloud. The server is also compared to the JCVI web server, which has its own versions of SIFT. The two databases differ in coverage, with the latest release including predictions for RefSeq, CCDS, and UCSC Known genes in addition to Ensembl gene annotations. The JCVI database is currently missing scores for chromosome Y.The SIFT web server predicts the effects of amino acid substitutions on protein function. First introduced in 2001, SIFT has become a standard tool for characterizing missense variation. The authors have updated the SIFT genome-wide prediction tool since their 2009 publication and added new features to the insertion/deletion (indel) tool. They also show accuracy metrics on independent data sets. The original developers host the SIFT web server at FHCRC, JCVI, and it is currently located at BII. The URL is http://sift-dna.org. SIFT uses sequence homology to compute the likelihood that an amino acid substitution will have an adverse effect on protein function. The algorithm was extended to predict the effects of frameshifting indels. SIFT has been used in human genetic research, including cancer, Mendelian diseases, and infectious diseases. It has also been used to study the effects of missense mutations in agricultural plants and model organisms. The SIFT algorithm was tested on the HumVar and HumDiv data sets, which were created by the authors of PolyPhen-2. The performance of SIFT was assessed using various protein databases, with UniRef90 chosen for its high coverage, sensitivity, and balanced performance. SIFT exhibited higher specificity on the HumDiv and HumVar data sets. New features of the web server include a genome-wide database of predictions for nonsynonymous variants and variant annotation with dbSNP and 1000 Genomes. The server also supports file format conversion, including VCF, Pileup, and GFF files. The indel prediction tool was developed to predict the effects of frameshifting indels, achieving 84% accuracy, 90% sensitivity, and 81% precision. The SIFT web server is freely available to the academic community. The source code and Linux executables are publicly available, and a public image is available on the Amazon cloud. The server is also compared to the JCVI web server, which has its own versions of SIFT. The two databases differ in coverage, with the latest release including predictions for RefSeq, CCDS, and UCSC Known genes in addition to Ensembl gene annotations. The JCVI database is currently missing scores for chromosome Y.