Sheng Wang

scWeave translates between gene expression and chromatin structure at single-cell resolution using deep learning, improving cross-modality inference across multiple datasets.

HiCFoundation is a Hi-C-based model that integrates chromatin structure with epigenomic regulation across species, improving analysis of 3D genome architecture and single-cell data.

SpineAgent generates spine MRI reports by integrating multi-sequence data using a multi-agent framework, achieving state-of-the-art performance and strong generalizability.

TwinC predicts trans-chromosomal DNA contacts using DNA sequence and provides insights into their functional mechanisms.

Puget predicts gene expression across cell types using sequence and Hi-C data, improving generalization to new biosamples and species.

Evo2HiC improves prediction of 3D genome architecture and epigenomic signals by integrating DNA sequences and Hi-C data, achieving better accuracy and computational efficiency.

LigUnity is a foundation model that jointly embeds ligands and protein pockets into a shared space to improve affinity prediction in virtual screening and hit-to-lead optimization.

The paper develops a pipeline for high-resolution imaging and analysis of whole mouse glomeruli, revealing novel aging and disease-related phenotypes.

Pisces enhances drug combination synergy prediction by generating 64 augmented views per drug pair, achieving state-of-the-art results across multiple prediction tasks.

CryoNeRF improves cryo-EM structure reconstruction by handling both homogeneous and heterogeneous data, achieving better performance than existing methods.

The paper shows autism genes converge at cilia, linking them to ciliary dysfunction and co-occurring conditions.

Capricorn enhances low-coverage Hi-C contact matrices using a multi-view diffusion model to improve loop detection accuracy across cell lines and chromosomes.

The paper constructs a PPI network of 100 high-confidence ASD genes, identifies novel interactions, and uses AlphaFold-Multimer to prioritize variants for functional validation in organoids.

This paper demonstrates how large language models can generate malicious abstracts to poison scientific knowledge graphs and manipulate downstream biological applications.

ActFound is a foundation model for bioactivity prediction trained on 2.3 million compounds and 50,869 assays, showing strong generalization and performance comparable to FEP+(OPLS4).

Sagittarius simulates gene expression at unmeasured timepoints using a transformer model, improving understanding of temporal dynamics in biology.

This paper introduces GraNA, a deep learning framework for supervised biological network alignment that improves functional annotation transfer between species.

Gemini integrates hundreds of gene networks using memory-efficient high-order pooling, improving protein function prediction performance.

scPretrain improves cell type classification by using self-supervised learning with both annotated and unannotated cells across 60 datasets.

This paper introduces a diffusion-based generative model for designing and optimizing antigen-specific antibodies by jointly modeling CDR sequences and structures.

The paper creates an organism-wide hormone signaling atlas in mouse lemurs, identifying hormone-producing and -target cells and revealing species-specific rewiring and conservation compared to humans and mice.

This paper introduces GraphPrompt, a graph-based prompt learning method for biomedical entity normalization, improving performance on zero-shot and few-shot settings using a curated dataset of entity-synonym pairs.

DrugOrchestra jointly predicts drug response, targets, and side effects using deep multi-task learning, showing improved performance over single-task methods.

cancerAlign improves tumor stratification by aligning tumors across cancer types using adversarial learning and consensus clustering, showing better results than methods that cluster within single cancer types.

MARS identifies and annotates known and novel cell types in single-cell RNA-seq data by transferring latent representations across datasets and generating interpretable names for new cell types.

The paper introduces scRFE, a method using random forests and recursive feature elimination to identify minimal gene sets distinguishing single-cell types in scRNA-seq data.

OnClass classifies single-cell data into Cell Ontology terms and identifies marker genes for all categories, even those absent in training data.

This paper introduces DeepSyn, a machine learning method for gene functional enrichment that uses a knowledge graph to infer gene functions based on disease and drug context, achieving high accuracy in various applications.

GRep represents gene sets as multivariate Gaussian distributions to improve drug target identification from expression-derived gene sets.