Research Pipeline

From raw tumor DNA to mRNA vaccine candidate — 5 autonomous steps powered by AI agents.

1
🔬

Mutation Discovery

Mutation Hunter

Scans tumor genomes from TCGA using ensemble variant calling (Mutect2 + Strelka2 + VarScan2) to identify somatic mutations.

Tools: GATK Mutect2Strelka2VarScan2Clara Parabricks
Compute: VAST.ai A100 or GDC API (pre-computed)
2
🎯

Neoantigen Screening

Neoantigen Scout

Predicts which mutations produce neoantigens visible to the immune system using peptide-MHC binding prediction.

Tools: pVACseqNetMHCpan 4.1MHCflurryNeoDesign
Compute: Railway CPU
3
🔮

Protein Structure

Structure Prophet

Predicts 3D structures of neoantigen peptide-MHC complexes to validate binding stability.

Tools: AlphaFold 3Boltz-2ESMFold
Compute: VAST.ai A100/L40S
4
🧬

mRNA Vaccine Design

mRNA Architect

Designs optimized mRNA sequences encoding selected neoantigens with maximum stability and immunogenicity.

Tools: LinearDesignmRNAidViennaRNA
Compute: Railway CPU
5
💊

Drug Simulation

Drug Simulator

Validates vaccine candidates through molecular docking and binding affinity simulations.

Tools: DiffDockRFdiffusion
Compute: VAST.ai L40S

CancelCancer — AI-powered mRNA cancer vaccine research. 100% open science.