Learning Latent Trajectories in Developmental Time Series with Hidden-Markov Optimal Transport

Peter Halmos, Julian Gold, Xinhao Liu, Benjamin J. Raphael

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Deriving the sequence of transitions between cell types, or differentiation events, that occur during organismal development is one of the fundamental challenges in developmental biology. Single-cell and spatial sequencing of samples from different developmental timepoints provide data to investigate differentiation but inferring a sequence of differentiation events requires: (1) finding trajectories, or ancestor:descendant relationships, between cells from consecutive timepoints; (2) coarse-graining these trajectories into a differentiation map, or collection of transitions between cell types, rather than individual cells. We introduce Hidden-Markov Optimal Transport (HM-OT), an algorithm that simultaneously groups cells into cell types and learns transitions between these cell types from single-cell or spatial developmental time series. HM-OT uses low-rank optimal transport to simultaneously align samples in a time series and learn a sequence of clusterings and a differentiation map with minimal total transport cost.

Original languageEnglish (US)
Title of host publicationResearch in Computational Molecular Biology - 29th International Conference, RECOMB 2025, Proceedings
EditorsSriram Sankararaman
PublisherSpringer Science and Business Media Deutschland GmbH
Pages367-370
Number of pages4
ISBN (Print)9783031902512
DOIs
StatePublished - 2025
Externally publishedYes
Event29th International Conference on Research in Computational Molecular Biology, RECOMB 2025 - Seoul, Korea, Republic of
Duration: Apr 26 2025Apr 29 2025

Publication series

NameLecture Notes in Computer Science
Volume15647 LNBI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference29th International Conference on Research in Computational Molecular Biology, RECOMB 2025
Country/TerritoryKorea, Republic of
CitySeoul
Period4/26/254/29/25

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Learning Latent Trajectories in Developmental Time Series with Hidden-Markov Optimal Transport'. Together they form a unique fingerprint.

Cite this