Bin Chen

This paper presents a drug repurposing framework using in vivo transcriptomics to expand indications by matching drug signatures with disease signatures, successfully recovering established uses of semaglutide and proposing new ones.

InsilicoCell predicts cellular functional profiles and enables multi-objective virtual drug screening by integrating diverse molecular data and tasks.

GEOMeta automates metadata curation for GEO transcriptomic data, generating standardized annotations for 600,000 human bulk RNA-seq samples and benchmarking LLMs for annotation quality.

This study identifies tumor-derived FXII as a driver of liver metastasis in breast cancer by engaging the intrinsic coagulation cascade, with anti-coagulation reducing metastasis in a mouse model.

This paper introduces RT-DMF, a deep learning framework to correct noise and impute missing values in pharmacogenomic datasets, improving downstream analysis.

The paper uses single-cell transcriptomics to identify S100 gene expression patterns and cell-type changes in MODS patients, showing diagnostic potential and therapeutic implications.

SPIDER predicts abundance of over 2500 surface proteins from single-cell transcriptomes using a context-agnostic zero-shot deep ensemble model, outperforming existing methods and enabling applications in cancer research.

The paper develops a computational pipeline to identify bispecific antibody targets by integrating bulk, single-cell, and spatial transcriptomics data, finding GPC3 and MUC13 as promising candidates in hepatocellular carcinoma.

The paper develops a tool to detect viral infections in cell lines using RNA-seq data and machine learning, finding 10% of MCF7 experiments were likely infected.

The paper identifies dihydroxyeicosadienoic acid as a metabolite of DGLA that induces ferroptosis-mediated neurodegeneration in dopaminergic neurons.

TransCell predicts genomic and cellular responses from gene expression using transfer learning, achieving strong performance in drug sensitivity and gene effect score prediction.

The paper identifies common mechanisms of action among anti-SARS-CoV-2 drugs and shows that targeting cholesterol homeostasis and microtubule cytoskeleton organization synergizes with antivirals.

OCTAD enables virtual screening of therapeutics for cancer patient groups using gene expression features, prioritizing potent drug candidates for experimental testing.

The paper evaluates cell lines as models of primary tumors across 22 types using transcriptomic data, finding cell-cycle pathways are consistently upregulated in cell lines.

This study evaluates the genomic similarity between breast cancer cell lines and metastatic samples, finding significant differences and highlighting organoids as better models.

The paper evaluates using GTEx normal tissue samples as references for cancer research and finds that a deep learning approach can effectively select appropriate normal samples for disease signature creation.

The paper uses longitudinal health records to identify a novel association between schizophrenia and rhabdomyolysis through analysis of hospital readmission patterns.

The paper introduces a framework to correct inconsistencies in pharmacogenomics datasets by improving drug-wise correlations while preserving cell-wise consistency.