retina-score-snps-q1

Question

https://zenodo.org/records/6330053 contains TensorFlow 2 models that are used for scoring the impact of SNPs on ATAC-seq in various cells types, as illustrated in the ScoreSNPs.ipynb notebook in https://zenodo.org/records/6796067. retina.score.snps.q1.tsv contains 500 SNPs (hg19 coordinates), of which half are known to be caQTLs. I would like to score the SNPs following the code in ScoreSNPs.ipynb, using the models corresponding to Rod cell type. After scoring the SNPs, report the AUPRC against the labels given in retina.score.snps.q1.tsv when using the absolute value of Rod_fold_avg_lfc as a predictor. Report answer by rounding down to two decimal places in the format: 0.80.

Inferred answer

0.67Uniquely inferred from submitted scores.

Domain

Machine Learning

Question style

Tooling

An answer can be uniquely inferred for 54 of 100 questions from submitted scores. Green: correct · Red: incorrect. For questions without a uniquely determined answer: Blue: matches prediction · Orange: differs from prediction. These are inferred answer checks, not official per-question grades.