ni-lab / personalized-expression-benchmark

Evaluating genomic sequence models for explaining personalized expression variation
MIT License
18 stars 2 forks source link

PEER Normalized Expression Labels #5

Open e-trop opened 10 months ago

e-trop commented 10 months ago

Hi there,

Thanks for this exciting work! I wanted to better understand the expression labels used in this study. How exactly were the labels from predictions/geuvadis_peer_normalized.csv processed - i.e. what types of transformations were applied to the original counts data?

Thank you!

xliu2525 commented 3 months ago

You can find the description, GeuvadisRNASeqAnalysisFiles_README.txt from the website, https://www.ebi.ac.uk/biostudies/arrayexpress/studies/E-GEUV-1. I think it's the same data as "GD462.GeneQuantRPKM.50FN.samplename.resk10.txt.gz".