EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A deep learning–based score to evaluate multiple sequence alignments

Nimrod Serok iD, Ksenia Polonsky iD, Haim Ashkenazy iD, Itay Mayrose iD, Jeffrey L Thorne iD, Tal Pupko iD

Molecular Biology and Evolution · 01.08.2026 · 43 (8)

DOI: 10.1093/molbev/msag173 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Molecular Biology and Evolution, online erschienen am 22.07.2026. Verfasst von Nimrod Serok und 5 weiteren Autor:innen. Beteiligte Einrichtung: The Shmunis School of Biomedicine and Cancer Research, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv 69978, Israel. Hauptthemen laut MEDLINE: Sequence Alignment, Deep Learning. Die Arbeit nennt 111 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Israel Science Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Multiple sequence alignment (MSA) inference is a central task in molecular evolution and comparative genomics, and the reliability of downstream analyses, including phylogenetic inference, depends critically on alignment quality. Despite this importance, most widely used MSA methods optimize the sum-of-pairs (SoP) score, and relatively little attention has been paid to whether this objective function accurately reflects alignment accuracy. Here, we evaluate the performance of the SoP score using simulated and empirical benchmark alignments. For each dataset, we compare alternative MSAs derived from the same unaligned sequences and quantify the relationship between their SoP scores and their distances from a reference alignment. We show that the alignment with the optimal SoP score often does not correspond to the most accurate alignment. To address this limitation, we develop deep learning–based scoring functions that integrate a collection of MSA features. We first introduce Model 1, a regression model that predicts the distance of a given MSA from the reference alignment. Across simulated and empirical datasets, this learned score correlates more strongly with true alignment accuracy than the SoP score. However, Model 1 is less effective at identifying the best alignment among alternatives. We therefore develop Model 2, which takes as input a set of alternative MSAs generated from the same sequences and predicts their relative ranking. Model 2 more accurately identifies the top-ranking MSA than the SoP score, Model 1, and several widely used alignment programs. Using simulations, we show that selecting MSAs based on our approach leads to more accurate phylogenetic reconstructions.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Sequence AlignmentDeep LearningPhylogenyAlgorithmsEvolution, Molecular

Stichwörter der Autor:innen

alignment algorithmsalignment scoring functionmachine-learningphylogenomicssum of pairs

Autor:innen und Institutionen

  • The Shmunis School of Biomedicine and Cancer Research, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv 69978, Israel.
  • Department of Molecular Biology, Max Planck Institute for Biology Tübingen, Tübingen, Germany.
  • The School of Plant Sciences and Food Security, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv 69978, Israel.
  • Department of Biological Sciences, North Carolina State University, Raleigh, NC 27695, USA.
  • Department of Statistics, North Carolina State University, Raleigh, NC 27695, USA.

ORCID-Profile: Nimrod Serok, Ksenia Polonsky, Haim Ashkenazy, Itay Mayrose, Jeffrey L Thorne, Tal Pupko

Förderung

  • Israel Science Foundation

Zitieren

APA

Serok, N., Polonsky, K., Ashkenazy, H., Mayrose, I., Thorne, J. L., & Pupko, T. (2026). A deep learning–based score to evaluate multiple sequence alignments. Molecular Biology and Evolution, 43(8). https://doi.org/10.1093/molbev/msag173

Vancouver

Serok N, Polonsky K, Ashkenazy H, Mayrose I, Thorne JL, Pupko T. A deep learning–based score to evaluate multiple sequence alignments. Molecular Biology and Evolution. 2026;43(8). doi:10.1093/molbev/msag173

BibTeX
@article{Serok2026,
  author = {Serok, Nimrod and Polonsky, Ksenia and Ashkenazy, Haim and Mayrose, Itay and Thorne, Jeffrey L and Pupko, Tal},
  title = {A deep learning–based score to evaluate multiple sequence alignments},
  journal = {Molecular Biology and Evolution},
  year = {2026},
  volume = {43},
  number = {8},
  doi = {10.1093/molbev/msag173}
}
Kostenlos starten →