Spaces:
Running
Running
File size: 1,605 Bytes
0a93d5d 021fcba 0a93d5d 021fcba b848be3 021fcba b848be3 3a49a77 0a93d5d b848be3 021fcba b848be3 021fcba 0a93d5d b848be3 0a93d5d b848be3 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 | # REAL COMPLETED EXAMPLE
# This file is based on the ESM-2 embedding processed in this
# project. Do not upload it unchanged. Copy submission.yaml and
# describe your own embedding. The ID esm2 is already in the catalogue.
schema_version: 1
embedding_id: esm2
display_name: ESM-2
contributor:
name: Daniel Tabares Lopez
affiliation: ETH Zurich
scientific:
category: Protein sequence
modality: Protein sequence
description: Protein sequence representation from ESM-2 650M, obtained by pooling residue-level embeddings into one vector per human protein/gene.
original_method: ESM-2 protein language model
model_checkpoint: esm2_t33_650M_UR50D
source_publication: https://doi.org/10.1101/2025.01.29.635607
provenance:
species: Homo sapiens
identifier_before_mapping: Entrez Gene ID
mapping_method: Source Entrez Gene IDs were mapped to the project's enriched human master gene table. Only identifiers mapping uniquely to one Ensembl gene were retained; unmapped and ambiguous identifiers were excluded, and duplicate source rows mapping to the same final gene were averaged.
embedding_generated_by: Jeffrey Zhong, Lechuan Li, Ruth Dannenfelser, and Vicky Yao
input_data_source: UniProt human protein sequences used in the Zhong et al. gene-embedding benchmark
pooling_strategy: Mean pooling over amino-acid token embeddings
licensing:
redistribution_allowed: true
attestations:
information_is_accurate: true
rights_to_redistribute: true
no_sensitive_human_data: true
no_patient_level_data: true
source_terms_reviewed: true
acceptance_of_automatic_checks: true
|