Sanju Sinha

This paper develops DECODEM to predict chemotherapy response by integrating deconvolved expression patterns of immune, stromal, and tumor cells in breast cancer, finding that cell-type-specific expression is predictive of response.

The paper introduces PathStAR, a framework to quantify structural aging from histopathology images, revealing tissue-specific aging trajectories and coordinated deterioration across organs.

PathQC predicts RNA Integrity Number and autolysis from H&E images using deep learning, achieving moderate to high correlations across tissues.

This paper introduces LogiCAR designer, a computational method using single-cell transcriptomics to identify logic-gated antigen combinations for CAR therapy, showing improved efficacy and safety compared to existing approaches.

TLPath predicts bulk-tissue telomere length from histopathology images, showing aging-related morphological changes and outperforming chronological age as a predictor.

The paper compares gene expression in primary and metastatic tumors to their origin and target tissues, finding metastases more resemble target tissues and show reduced cancer hallmark activity.

ecPath detects ecDNA in tumors using H&E images, achieving good performance in predicting ecDNA status and identifying poor prognoses in low grade glioma patients.

IMMClock predicts immune age of T cells and NK cells using gene expression, showing immune aging is more linked to cellular aging than differentiation and is accelerated in diseases.

The paper infers tumor immune status and predicts immunotherapy response from blood scRNA-Seq data in HNSCC patients.

PROTGOAT improves protein function prediction using an ensemble of pre-trained protein language models, achieving high performance in CAFA 5 benchmark.

ENLIGHT predicts patient response to targeted and immunotherapies using tumor transcriptomes without prior treatment response data, showing better accuracy than existing transcriptomics-based signatures.

The paper predicts anti-SARS-CoV-2 targets by identifying synthetic lethal partners of host genes altered during infection, validated through CRISPR-Cas9 screens and siRNA experiments.

This paper introduces PERCEPTION, a framework using single-cell transcriptomics to predict cancer treatment response and resistance, outperforming bulk expression-based methods in clinical trials.

The paper uses genome-scale metabolic modeling to identify SARS-CoV-2-induced metabolic changes and predict antiviral targets validated through drug and genetic screens.

The paper analyzes pan-cancer single-cell transcriptomes to evaluate and identify new CAR targets for head and neck cancer, finding five promising targets with superior selectivity and safety.

The paper uses single-cell transcriptomics data to explore optimal combination therapies for cancer by analyzing receptor targets across multiple tumor types.

The paper compares genomic features between African American and European American lung cancers, finding higher genomic instability and homologous recombination-deficiency in African American tumors.