Enhanced Detection of Homology Using Artificial Intelligence in Euglenids

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Identification of similarity between protein sequences is an important component for the assignment of function. With ever-growing databases of genome sequence, this becomes an increasing challenge, and especially in the detection of relationships between distantly related sequences, which is frequently an issue with euglenids. The introduction of artificial intelligence tools to the prediction of protein structure has been, without exaggeration, revolutionary. In particular, AlphaFold3 (AF3), the latest iteration of the AI predictor from DeepMind, a Google subsidiary, offers a potent combination of speed, accuracy, and ease-of-use, all free of charge. Here I will describe a basic workflow for the detection of low similarity between proteins, that is otherwise cryptic, using AF3, discuss how to interpret the predictions, and highlight examples of bizarre predictions or hallucinations.

Original languageEnglish
Title of host publicationEuglenozoa
Subtitle of host publicationMethods and protocols
EditorsPaul A M Michels, Michael L. Ginger, Anna Karnkowska, Laura-Isobel McCall, Ariel M Silber
Place of PublicationNew York
PublisherHumana Press
Pages401-409
Number of pages9
Volume1
ISBN (Electronic)9781071651421
ISBN (Print)9781071651445, 9781071651414
DOIs
Publication statusPublished - 3 Feb 2026

Publication series

NameMethods in Molecular Biology
Volume3013
ISSN (Print)1064-3745
ISSN (Electronic)1940-6029

Keywords

  • AI
  • AlphaFold
  • Homology
  • Protein structure
  • Sequence evolution

ASJC Scopus subject areas

  • Molecular Biology
  • Genetics

Fingerprint

Dive into the research topics of 'Enhanced Detection of Homology Using Artificial Intelligence in Euglenids'. Together they form a unique fingerprint.

Cite this