google-deepmind / google-deepmind/deepmind-research

Meaning of EnformerScoreVariants output

Open
#373 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Jupyter Notebook
Stars
15.2k
Forks
2.9k
PR merge metrics
No merged PRs in 30d

Description

What do the Z-scores of EnformerScoreVariantsNormalized() represent? Is this the relative change in the signal over 200KB centered on the variant? When I find a large Z-score and then try to visualize this region, the differences don't seem to line up.

Also, what do the principle components from EnformerScoreVariantsPCANormalized() represent? Is this taking the relative differences amongst all the human tracks (5K dimensions) and reducing this into 20 dimensions? Is this so mutations with similar effects can be clustered? Are we able to tell effect size with this output?

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.