Wenjing Ma

Across 13 single-cell foundation models and 94 datasets, PCA and established methods usually match or beat them; pretrained embeddings help mainly with scarce labels, rare cells and open-set detection.

FastCCC provides a permutation-free framework for reference-based cell-cell communication analysis, demonstrating robust performance across complex datasets and uncovering biologically relevant interactions.

SpaDOT identifies spatial domains and their temporal dynamics in spatiotemporal transcriptomics data, revealing key developmental processes like valvulogenesis.

scCAFE is a deep learning model that detects multi-scale chromatin architectural features in single-cell Hi-C data, outperforming existing methods and enabling cell-type annotation from 3D genome features.

LRcell identifies sub-cell types responsible for differential expression in bulk RNA-seq by using precomputed marker genes from scRNA-seq data, outperforming existing deconvolution methods.

SELMA estimates intrinsic cleavage biases in bulk and single-cell chromatin accessibility data, improving analysis accuracy and cell clustering in single-cell ATAC-seq.

BARTweb identifies transcription factors associated with genomic features using public ChIP-seq data, outperforming existing tools and reducing computation time.