Tianwei Yu

The paper introduces a knowledge-driven framework that models biological interactions in omics data using pathway graphs, enabling interpretable mechanistic insights into disease progression.

NetworkSSC improves single-cell clustering by integrating gene network information, outperforming traditional methods on multiple scRNA-seq datasets.

NicheScope identifies multicellular niches and their regulated cell states from spatial transcriptomics data, revealing tissue structures and clinically relevant tumor niches.

This paper introduces IBSEP, a statistical framework that integrates bulk and scRNA-seq data to prioritize cell-type-specific eQTLs, showing improved performance over existing methods.

fast-scBatch corrects batch effects in scRNA-seq data using a neural network-driven approach, outperforming existing methods in simulation and real datasets.

ADM combines multiple dimension reduction techniques using graph diffusion to better capture complex data structures and sample heterogeneity.

The paper introduces a statistical method to identify spatially associated biological pathways, finding significant pathways linked to spatial variation in human and mouse datasets.

G3DC improves single-cell RNA-seq clustering by integrating gene networks and addressing data sparsity, achieving higher accuracy than existing methods.

The paper introduces DCOL-PCA and DCOL-CCA for nonlinear dimension reduction and integration of single- and multi-omics data, showing superior performance in simulations and real datasets.

FAST improves spatial dimension reduction for multi-section SRT data by modeling count data and spatial smoothness, showing better performance and speed on real datasets.

This study maps the molecular lineage of chordate embryogenesis using single-cell transcriptomics in Ciona savignyi, identifying 47 cell types and revealing fate determination mechanisms.

scBatch corrects RNA-seq batch effects by adjusting the sample distance matrix, improving clustering and differential expression results compared to existing methods.