Jinru Shia

This paper identifies a five-protein stem cell signature using a deep neural network that predicts relapse in stage III colorectal cancer patients.

The paper develops Mosaic, a machine learning system that classifies tumors into detailed subtypes and identifies molecular phenotypes using histopathologic and genomic data from a large patient cohort.

This study identifies genetic and non-genetic resistance mechanisms to KRAS inhibition in CRC using single-cell spatial transcriptomics and exome sequencing, revealing inflammatory and fetal-like transcriptional signatures associated with resistance.

The paper develops patient-derived organoids and mouse models of appendiceal cancer to identify phenotypic changes and drug targets in peritoneal metastasis.

The paper investigates the spatial relationship between T cells and cancer cells in stage III CRC patients, finding that closer proximity of T cells to cancer cells is associated with better disease-free survival.

MiMSI improves MSI detection in low tumor purity samples using deep learning and multiple instance learning, achieving higher sensitivity and auROC than existing methods.