Computational biology projects built with AI
Onco Strat
Ranks drugs for a breast cancer patient from their tumour's DNA and RNA
Built by Anshiika Bisht, Shashidhar K and Jagriti
What it does
In triple-negative breast cancer, trying one drug after another costs time the patient may not have. Onco Strat takes a tumour's variant file and its RNA-seq counts, keeps the variants that matter, and checks which of those genes are also switched on. It matches the result against OncoKB and produces a PDF report of candidate drugs, ranked by the AMP/ASCO/CAP evidence tiers.
How it works
- Filters the VCF to protein-changing, high or moderate impact, somatic variants on canonical transcripts
- Keeps only genes that are both mutated and upregulated in the RNA-seq data
- Matches them against OncoKB and writes a patient report with drugs and evidence tiers
Built with VCF · RNA-seq · Ensembl VEP · gnomAD · OncoKB · Python · Streamlit
A learner prototype built on public data, shared with the team's permission as part of their portfolio. It is not a clinical tool.
More projects
Oncomap
Turns a pancreatic cancer VCF into a molecular oncology report
See the project
Onco Compass
Finds the actionable mutations in a lung cancer sample
See the project
EvidOnco
An assistant that answers biomarker questions with sources
Ashish Balan, Nandini Solanki, Harsha S. and Anisha
See the project
Build your own in the PGP
Six months, live classes at the weekend, and a capstone project you can show employers.