I am a clinician and bioinformatics learner building practical tools for cancer multi-omics, single-cell analysis, biomedical foundation models, and AI-assisted research.
Clinical Medicine Single-cell Omics Biomedical AI
Cancer Multi-omics Foundation Models Deep Learning
TCGA / GDC Data Research Agents Reproducible Science
|
A fine-tuned Cell2Sentence Gemma-2B model for ulcerative colitis single-cell RNA-seq cell type prediction. Training data: 311,920 human colon cells, 11,743 genes, 44 L3 cell types from HC and UC samples. Evaluation: 89.00% accuracy on 500 test samples, a 25x improvement over the original pre-trained model on the UC dataset. |
|
Screen TCGA/GDC cohorts for complete multi-modal patient data, aligning RNA, CNV, WSI, and survival metadata with gdc-client manifests. Keywords: TCGA, GDC, multi-modal cancer data, WSI, survival analysis |
A Python toolkit for gene-centered biomedical interpretation and narrative generation. Keywords: gene interpretation, biomedical knowledge, Python |
|
A workflow for prioritizing candidate genes from biomedical evidence and analysis results. Keywords: gene prioritization, evidence ranking, bioinformatics |
AI-agent experiments for biomedical reading, research assistance, and workflow automation. Keywords: AI agents, research automation, biomedical reading |
