scHiGex: predicting single-cell gene expression based on single-cell Hi-C data
A graph transformer that predicts a single cell's gene expression directly from that same cell's 3D genome structure.
Single-cell Hi-C gene-expression prediction
Abstract
The HiRES experiment captures both chromosomal conformation and gene expression of individual single cells simultaneously, but such datasets remain limited compared to the volume of single-cell Hi-C data available. scHiGex is a graph transformer that predicts gene expression levels from single-cell Hi-C data. Benchmarking demonstrates strong performance with an average absolute error of 0.07, and the predicted expression levels yield precise cell-type categorization (adjusted Rand index of 1), showing the model captures heterogeneity between individual cell types. scHiGex is freely available on GitHub.
Authors
Shrestha B, Siciliano AJ, Zhu H, Liu T, Wang Z
NAR Genomics and Bioinformatics · 7(1), lqaf002 · Oxford University Press
How to cite
@article{shrestha2025schigex,
title = {scHiGex: predicting single-cell gene expression based on single-cell Hi-C data},
author = {Shrestha, Bishal and Siciliano, A. J. and Zhu, H. and Liu, T. and Wang, Zheng},
journal = {NAR Genomics and Bioinformatics},
volume = {7},
number = {1},
pages = {lqaf002},
year = {2025},
doi = {10.1093/nargab/lqaf002}
}