Posts by Collection

Portfolio item number 1

Short description of portfolio item number 1

Portfolio item number 2

Short description of portfolio item number 2

Contrastive Cycle Adversarial Autoencoders for Single-cell Multi-omics Alignment and Integration

Published in Bioinformatics, 2021

Xuesong Wang, Zhihang Hu, Tingyang Yu, Ruijie Wang, Yumeng Wei, Juan Shu, Jianzhu Ma, Yu Li

Muilti-modality data are ubiquitous in biology, especially that we have entered the multi-omics era, when we can measure the same biological object (cell) from different aspects (omics) to provide a more comprehensive insight into the cellular system. When dealing with such multi-omics data, the first step is to determine the correspondence among different modalities. In other words, we should match data from different spaces corresponding to the same object. To promote the single-cell multi-omics research, we proposing a novel framework to align and integrate single-cell RNA-seq data and single-cell ATAC-seq data. Our approach can efficiently map the above data with high sparsity and noise from different spaces to a low-dimensional manifold in a unified space, making the downstream alignment and integration straightforward. Compared with the other state-of-the-art methods, our method performs better in both simulated and real single-cell data. The proposed method is helpful for the single-cell multi-omics research. The improvement for integration on the simulated data is significant.

conST: an interpretable multi-modal contrastive learning framework for spatial transcriptomics

Published in biorXiv (Major revision at Nucleic Acids Research), 2022

Yongshuo Zong, Tingyang Yu, Xuesong Wang, Yixuan Wang, Zhihang Hu, Yu Li

Spatially resolved transcriptomics (SRT) shows its impressive power in yielding biological insights into neuroscience, disease study, and even plant biology. We propose conST, a powerful and flexible SRT data analysis framework utilizing contrastive learning techniques. conST can learn low-dimensional embeddings by effectively integrating multi-modal SRT data, i.e. gene expression, spatial information, and morphology (if applicable). The learned embeddings can be then used for various downstream tasks, including clustering, trajectory and pseudotime inference, cell-to-cell interaction, etc. Our framework is interpretable in that it is able to find the correlated spots that support the clustering, which matches the CCI interaction pairs as well, providing more confidence to clinicians when making clinical decisions.’

scMinerva: an Unsupervised Graph Learning Framework with Label-efficient Fine-tuning for Single-cell Multi-omics Integrative Analysis

Published in biorXiv, 2022

Tingyang Yu, Yongshuo Zong, Yixuan Wang, Xuesong Wang, Yu Li

Single-cell multi-omics is a rapidly growing field in biomedicine, where multiple biological contents, such as the epigenome, genome, and transcriptome, can be measured simultaneously. Despite its potential, the integrated analysis and prediction of cellular states based on this complex multi-omics data pose significant challenges due to data sparsity, high noise, and computational overhead. To address these challenges, we developed scMinerva, an unsupervised framework for single-cell multi-omics integrated analysis. The learned embeddings from the multi-omics data enable accurate integrated classification of cell types and stages. Specifically, we construct a heterogeneous graph from multiple omics and propose a novel biased random walk algorithm omics2vec, which can learn the heterogeneous biological graph in a way that balances both local and global network structures. scMinerva successfully outperforms existing unsupervised methods on various simulated and real-world datasets when fine-tuned by very few labels.

Projection Robust Optimal Transport Between Unbalanced Distributions

Published in preprint, 2022

Tingyang Yu*, Yuxuan Wan*, Shiqian Ma

We propose a novel unbalanced optimal transport (UOT) formulation that has the potential to alleviate the curse of dimensionality. The key idea is borrowed from recent developments of the projection robust Wasserstein distance, which projects the sampled data onto lowerdimensional subspace and computes the Wasserstein distance between the projected data. Using the same idea, we propose the projection robust UOT, which is a max-min problem over Stiefel manifold. We propose two algorithms for solving this problem and analyze their complexity for obtaining an ϵ-stationary point. Numerical experiments on both synthetic and real datasets are conducted to demonstrate the advantages of our new UOT formulation in high-dimensional cases.