scTopoGAN: unsupervised manifold alignment of single-cell data

Akash Singh; Kirti Biharie; Marcel J T Reinders; Ahmed Mahfouz; Tamim Abdelaal

doi:10.1093/bioadv/vbad171

scTopoGAN: unsupervised manifold alignment of single-cell data

Bioinform Adv. 2023 Nov 24;3(1):vbad171. doi: 10.1093/bioadv/vbad171. eCollection 2023.

Authors

Akash Singh¹, Kirti Biharie^{1

2

3}, Marcel J T Reinders^{1

2

3}, Ahmed Mahfouz^{1

2

3}, Tamim Abdelaal^{1

2

4}

Affiliations

¹ Delft Bioinformatics Lab, Delft University of Technology, 2628 XE Delft, The Netherlands.
² Leiden Computational Biology Center, Leiden University Medical Center, 2333 ZC Leiden, The Netherlands.
³ Department of Human Genetics, Leiden University Medical Center, 2333 ZC Leiden, The Netherlands.
⁴ Department of Radiology, Leiden University Medical Center, 2333 ZC Leiden, The Netherlands.

Abstract

Motivation: Single-cell technologies allow deep characterization of different molecular aspects of cells. Integrating these modalities provides a comprehensive view of cellular identity. Current integration methods rely on overlapping features or cells to link datasets measuring different modalities, limiting their application to experiments where different molecular layers are profiled in different subsets of cells.

Results: We present scTopoGAN, a method for unsupervised manifold alignment of single-cell datasets with non-overlapping cells or features. We use topological autoencoders (topoAE) to obtain latent representations of each modality separately. A topology-guided Generative Adversarial Network then aligns these latent representations into a common space. We show that scTopoGAN outperforms state-of-the-art manifold alignment methods in complete unsupervised settings. Interestingly, the topoAE for individual modalities also showed better performance in preserving the original structure of the data in the low-dimensional representations when compared to other manifold projection methods. Taken together, we show that the concept of topology preservation might be a powerful tool to align multiple single modality datasets, unleashing the potential of multi-omic interpretations of cells.

Availability and implementation: Implementation available on GitHub (https://github.com/AkashCiel/scTopoGAN). All datasets used in this study are publicly available.