st2traj: deconvolution-informed trajectory inference for multi-timepoint spatial transcriptomics.

Authors

Zhuo Wang, Chiping Zhang

Published in

Bioinformatics (Oxford, England). Aug 27, 2026. Epub Aug 27, 2026.

Abstract

Multi-timepoint spatial transcriptomics enables study of developmental processes in native tissue context, but cell-state mixtures within spots and lack of direct spatial correspondence across sections complicate trajectory inference and biological interpretation.
st2traj is a deconvolution-informed trajectory framework using spot-state composition for multi-timepoint spatial trajectory inference. In a human heart pseudo-spot benchmark, DECODE showed competitive and balanced performance among five deconvolution methods. In multi-timepoint human heart data, unscaled DECODE-derived proportions produced smoother trajectory fields and stronger agreement with expression-derived marker programs than normalized spot-level expression. st2traj also showed greater spatial coherence than spaTrack, while exploratory comparisons with moscot and CASCAT revealed complementary method-specific strengths. Application to an independent chicken heart dataset recovered stage-associated trajectory changes across D7, D10, and D14.
Source code: https://github.com/xiaoxiaoxier/st2traj. Software v0.1.0 and processed data are archived at Zenodo: https://doi.org/10.5281/zenodo.21487030 and https://doi.org/10.5281/zenodo.21502094.
Supplementary data are available at Bioinformatics online.

PMID:
42658029
Bibliographic data and abstract were imported from PubMed on 27 Aug 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.