metadata

license: cc
task_categories:
  - sentence-similarity
tags:
  - biology
  - protein

Protein Pairs and Similarity

Selected protein similarities within training, test, and validation sets. Each protein gets two similarities selected at random and (usually) proteins within the top and bottom quintiles for similarity.

The protein is represented by its UniProt ID and its amino acid sequence (using IUPAC-IUB codes where each amino acid maps to a letter of the alphabet, see: https://en.wikipedia.org/wiki/FASTA_format ).

The distance column is cosine distance (identical = 0) on UniProt / SwissProt 1,024-dimension embeddings, downloaded in March 2024, based on the https://huggingface.co/Rostlab/prot_t5_xl_uniref50 model.

For models and training scripts optimizing for similarity, you should use (1 - distance).

Train / test / validation splits are from khairi/uniprot-swissprot

Code notebook: https://colab.research.google.com/drive/1rhoF0pAauHbiaNHdZkOI5O7A3mMcvS6G?usp=sharing

Source of proteins: UniProt-SwissProt

Download 1,024-dim embeddings from UniProt: https://www.uniprot.org/help/downloads#embeddings