dbnsfp v3.0 damaging prediction

Question

[kvshamsudheen]

I was using the dbnsfp v3.0 database for annotating a large vcf dataset of WES. I have difficulties in determining the damaging prediction of variants especially the one from muti-transcript genes.

Is there any way to determine the pathogenicity based on the most relevant prediction, about a variant based on canonical/noncanonical annotations (assuming that one variant has both damaging, probably damaging and neutral/benign prediction from multiple transcripts).

One way is to use the converted rank score, but that won't help in deciding the transcript is relevant or not

Answer 1

[Kevin Blighe]

People generally choose the canonical transcript, as it is assumed to be the one that is most expressed. For the majority of transcript isoforms, however, we just don't have the information about which isoforms are expressed in which tissues and under which conditions. Note also that people's interpretation of canonical differs. To ensembl, the canonical is merely the longest transcript, whereas to others it is the one that is most highly expressed in the tissue of interest. See here: https://www.ensembl.org/Help/Glossary?id=346

Regarding variant filtering, generally, there are no specific standards in place, but research is active in this area currently in both academic and private sectors. It is a core interest of mine.

For the US perspective on what is important for variant filtering, please read:

• Standards and Guidelines for the Interpretation and Reporting of Sequence Variants in Cancer: A Joint Consensus Recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists

You should also be aware of the work of Heidi Rehm, who has been a leading figure in this area in the USA (anybody who has worked in clinical genetics will have heard of her).

For the British perspective, it's best to read the material by the NGRL in Manchester, UK:

• http://www.ngrl.org.uk/Manchester/publications

Kevin