SilVA: Silent Variant Analysis using random forests.

About

SilVA (Latin for "forest") is a tool for the automated harmfulness prediction of synonymous (silent) mutations within the human genome. SilVA bases its predictions on a number of features, including conservation, codon usage, splice sites, splicing enhancers and suppressors, and mRNA folding free energy. Given variants in a VCF file, SilVA will rank the rare synonymous variants according to their predicted harmfulness. On a single machine with multiple CPUs, it takes us about 10 minutes to run SilVA (start to finish) on the ~10,000 synonymous SNVs in a typical exome sequencing experiment.

SilVA was developed at the University of Toronto Computational Biology Lab. Please see the latest README for information regarding installation and usage.

News

12 July 2013: Version 1.1.1 released. Updated to use hg19.2bit from UCSC in order to reduce demand on cs.toronto.edu servers.

11 July 2013: Released an alternative version of SilVA without UNAfold and ViennaRNA to simplify the installation for those who do not need those features. As mentioned in the paper, we did not find these features to substantially affect the performance, so this should not be a problem for most users. The link is directly below the regular download button.

10 July 2013: Updated ViennaRNA download URL in SilVA v1.1.0 source.

4 June 2013: Published online at doi:10.1093/bioinformatics/btt308.

24 April 2013: Version 1.1.0 released. Streamlined pipeline, added ViennaRNA features, improved input and output formats.

30 September 2012: Version 1.0.1 released. Dependencies reduced, output format changed, and other minor changes in preparation for reviewiers.

13 May 2012: Website online. Please note that the website and public code are still under development. Thank you for your patience.

Citatation

Buske OJ, Manickaraj A, Mital S, Ray PN, Brudno M. (2013) Identification of deleterious synonymous variants in human genomes. Bioinformatics, doi:10.1093/bioinformatics/btt308