Hua Tan

SoMAS identifies somatic mutations linked to alternative splicing in 33 cancer types, revealing 908 genes associated with altered isoform expression across multiple cancers.

This study develops a 2YS logistic regression model using radiogenomics analysis to predict GBM patient survival based on MRI features and gene associations.

The paper quantifies gene and isoform expression changes in metastatic melanoma compared to primary tumors, identifies distinct expression patterns linked to mutational subtypes, and highlights potential therapeutic targets.

This study introduces FusionAI, a deep learning model that predicts gene fusion breakpoints from DNA sequences, aiding in the identification of genomic breakage mechanisms.

The paper analyzes miRNA-gene positive correlations across 31 cancers using TCGA data, finding their prevalence, clinical relevance, and potential epigenetic mechanisms.