Tianyu Wu

PhD

PhD Student

Biostatistical Programming Data Mining multi-omics analysis (Transcriptome/ATAC/ChIP/Methylome/Proteome/Metabolome/WES/Single-cell multiome) Neuroimaging Analysis Python R Programming

1st author peer-reviewed publication(s) Conference Talks(s) Reviewed Publication(s)

I developed a machine learning framework that leverages whole-brain functional connectivity profiling to dissect biological heterogeneity in obesity. I designed and implemented unsupervised clustering algorithms to identify neurofunctionally distinct subtypes, then established prospective validation strategies to test whether these brain-based subgroups differentially predict therapeutic outcomes to metabolic interventions. This work bridges computational neuroimaging and precision medicine, providing a data-driven approach to patient stratification in metabolic disease management.

I lead a multi-omics investigation to uncover biomarkers of cognitive impairment in type 2 diabetes, integrating genomic, proteomic, neuroimaging, and clinical data through a unified analytical pipeline that I built from scratch. I developed algorithms for deep quantitative proteomic profiling and predictive modeling to distinguish diagnostic and prognostic signatures. The project spans from raw data processing and statistical validation to biological interpretation, translating multi-dimensional disease characteristics into clinically actionable frameworks.

I independently designed and deployed a centralized data infrastructure for a large-scale longitudinal cohort study, building automated pipelines for questionnaire collection, real-time data validation, and integrated neuroimaging analysis. I architected the backend using functional programming principles and containerized deployment strategies to ensure scalability and reproducibility. This work demonstrates my ability to bridge complex clinical research needs with robust software engineering, streamlining data acquisition through to analysis.