Datasets were updated on 2017-02-16, see release notes.

VariSNP is a benchmark database suite comprising variation datasets that can be used for developing and testing the performance of variant effect prediction tools.  VariSNP contains datasets selected from dbSNP which were filtered for disease-related variants found in ClinVar, Swiss-Prot and PhenCode, so all variations are considered neutral/non-pathogenic. The datasets are categorized according to the functional classes used in dbSNP.

Currently VariSNP contains the following datasets:

• Neutral Single Nucleotide Variants
  446 013 single nucleotide variants which alter the codon to make an altered amino acid in the protein product.
• Neutral non-coding transcript variants
  956 958 transcript variants of a non-coding RNA gene (located within a gene that does not code for a protein).
• Neutral downstream variants 500B
  470 473 sequence variants located within a half kb of the end of a gene.
• Neutral frameshift variants
  3802 sequence variants which cause a disruption of the translational reading frame, because the number of nucleotides inserted or deleted is not a multiple of three.
• Neutral cds indel variants
  9285 indel variants with length of multiple of 3 bp, thus not causing frameshift.
• Neutral splice acceptor variants
  3402 splice variants that change the 2 base region at the 3' end of introns.
• Neutral splice donor variants
  5277 splice variants that change the 2 base pair region at the 5' end of introns.
• Neutral stop gained variants
  11 339 sequence variants whereby at least one base of a codon is changed, resulting in a premature stop codon and leading to a shortened transcript.
• Neutral stop lost variants
  588 sequence variants where at least one base of the terminator stop codon is changed, resulting in an elongated transcript.
• Neutral synonymous codon variants
  318 967 sequence variants where there is no resulting change to the encoded amino acid.
• Neutral upstream variants 2KB
  1 804 501 sequence variants located within 2 kb 5' of a gene
• Neutral utr-3-prime variants
  610 396 variants of the 3' UTR
• Neutral intron variants
  25 930 776 intronic variants

An explanation of the columns in these datasets can be found in Legends of columns

The current datasets were generated and uploaded on 2017-02-16, dbSNP update builds 149, GRCh38p7. The old datasets can be found in the old directory.
See also the release notes.