EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

NEAR: neural embeddings for amino acid relationships

Daniel Olson, Thomas Colligan, Daphne Demekas, Jack W Roddy, Ken Youens-Clark, Travis J Wheeler iD

Bioinformatics · 2025

DOI: 10.1093/bioinformatics/btaf198 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bioinformatics, 2025. Verfasst von Daniel Olson und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, University of Montana, Missoula, MT 59812, United States. Hauptthemen laut MEDLINE: Sequence Alignment, Neural Networks, Computer, Proteins, Amino Acids. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 58 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIH, National Science Foundation, NIGMS NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Summary Protein language models (PLMs) have recently demonstrated potential to supplant classical protein database search methods based on sequence alignment, but are slower than common alignment-based tools and appear to be prone to a high rate of false labeling. Here, we present Neural Embeddings for Amino acid Relationships (NEAR), a method based on neural representation learning that is designed to improve both speed and accuracy of search for likely homologs in a large protein sequence database. NEAR’s ResNet embedding model is trained using contrastive learning guided by trusted sequence alignments. It computes per-residue embeddings for target and query protein sequences, and identifies alignment candidates with a pipeline consisting of residue-level k-NN search and a simple neighbor aggregation scheme. Tests on a benchmark consisting of trusted remote homologs and randomly shuffled decoy sequences reveal that NEAR substantially improves accuracy relative to state-of-the-art PLMs, with lower memory requirements and faster embedding and search speed. While these results suggest that the NEAR model may be useful for standalone homology detection with increased sensitivity over standard alignment-based methods, in this manuscript, we focus on a more straightforward analysis of the model’s value as a high-speed pre-filter for sensitive annotation. In that context, NEAR is at least 5x faster than the pre-filter currently used in the widely used profile hidden Markov model (pHMM) search tool HMMER3, and also outperforms the pre-filter used in our fast pHMM tool, nail. Availability and implementation NEAR is under an open-source license. Code and data curation instructions can be found at https://github.com/TravisWheelerLab/NEAR.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Sequence AlignmentNeural Networks, ComputerProteinsAmino AcidsSequence Analysis, ProteinDatabases, ProteinAlgorithmsAmino Acid Sequence

Substanzen

ProteinsAmino Acids

Autor:innen und Institutionen

  • Department of Computer Science, University of Montana, Missoula, MT 59812, United States.
  • College of Pharmacy, University of Arizona, Tucson, AZ 85721, United States.

ORCID-Profile: Travis J Wheeler

Förderung

  • NIH
  • National Science Foundation
  • NIGMS NIH HHS
  • Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support
  • Jetstream2
  • NIGMS

Zitieren

APA

Olson, D., Colligan, T., Demekas, D., Roddy, J. W., Youens-Clark, K., & Wheeler, T. J. (2025). NEAR: neural embeddings for amino acid relationships. Bioinformatics. https://doi.org/10.1093/bioinformatics/btaf198

Vancouver

Olson D, Colligan T, Demekas D, Roddy JW, Youens-Clark K, Wheeler TJ. NEAR: neural embeddings for amino acid relationships. Bioinformatics. 2025. doi:10.1093/bioinformatics/btaf198

BibTeX
@article{Olson2025,
  author = {Olson, Daniel and Colligan, Thomas and Demekas, Daphne and Roddy, Jack W and Youens-Clark, Ken and Wheeler, Travis J},
  title = {NEAR: neural embeddings for amino acid relationships},
  journal = {Bioinformatics},
  year = {2025},
  doi = {10.1093/bioinformatics/btaf198}
}
Kostenlos starten →