broadinstitute/SpliceAI-lookup

TET2 Cannonical transcript not selected

Closed this issue · 3 comments

https://spliceailookup.broadinstitute.org/#variant=chr4-106162586-G-T&hg=37&distance=50&mask=0&precomputed=0URL (copy-paste the url of a page where you see the issue or have a feature suggestion):

The score is computed selecting Use Illumina's pre-computed scores: yes, result is:

Screenshot (optionally, paste or drag-drop an image here):
image

However, when selecting

The score is computed selecting Use Illumina's pre-computed scores: no, result is:

image

and the highest score is the canonical transcript with the highest score but it is not consistent when using yes. Why is the difference?

Thanks,

Is this team still supporting spliceAI?

bw2 commented

Hi @cpsosa0006 , the Illumina pre-computed scores are based on gencode v26, while the other scores are based on gencode v42 - so that's the likely reason.

bw2 commented

I'm not part of the SpliceAI team. I just made this web server to look up or compute SpliceAI scores developed @ Illumina.