Zexian Liu

Synora identifies tumor-stroma boundaries in spatial omics data using cell coordinates and tumor/non-tumor labels, achieving robust boundary detection across datasets and modalities.

The paper introduces iHypoxia, a database compiling protein expression dynamics under hypoxia across mammals, revealing enriched drug targets and PTMs.

dSCOPE predicts sequence segments critical for liquid-liquid phase separation (SCOPEs) using a random forest model trained on curated experimental data, showing enrichment in post-translational modifications and cancer mutations.