synVep

What do the columns mean?

Gene: HGNC gene symbol
Ensembl: Ensembl transcript ID
Chr: chromosome
Position_hg19: genomic position (GRCh37/hg19)
Ref: reference allele (positive strand)
Alt: alternative allele (positive strand)
dbSNP: dbSNP ID
Class: "observed", "not-seen", "singletons", and "unobservable".
"observed" and "not-seen" are the labels used in the model development;
predictions are not meaningful if the variant class is "unobservable".
synVep: synVep-predicted score
Effect: synVep binary prediction: effect or no-effect
Position transcript: position of the variant allele on the transcript
Strand: the coding DNA strand
Codon: variant-induced synonymous codon change
Position_hg38: genomic position (GRCh38/hg38)

Why the database does not return predictions for my variants?

It could be multiple reasons. Are your variants synonymous? Are they located on transcripts that do not pass our quality control filters (see Methods in the synVep paper)? Here is a list of transcript IDs that are filtered out from our data Is the format correct? If you checked all of these and still do not have an answer, please contact us

Why some non-synonymous variants are returned as results?

The same variant could lead to different consequences (synonymous, missense, etc.) due to multiple transcripts that the variant hits.

Will my data be accessible to others?

We do not store or distribute your data.

How to query data locally?

We recommend using our local version of query: users need to download a Python script (here) and the database (here). The Python (version 3) script is simple to use and has no dependency.