ching-goodfire/

MAPS-ClinVar-VKS-Inputs

Two normalized tables that carry the full information content of the original 406.8 MB harvest_manifest.parquet in 9.8 MB, plus the ClinVar identifiers needed to trace any variant back to its source record.

	Files




	

file rows contents

variants.parquet 200,913 one row per variant: identity, split, label, cluster, ClinVar ids

wt_sequences.parquet 17,231 one row per RefSeq protein: protein_id… See the full description on the dataset page: https://huggingface.co/datasets/ching-goodfire/MAPS-ClinVar-VKS-Inputs.