Upload a tumor biopsy file. Get ranked vaccine targets, an AI clinical report, and a synthesis-ready mRNA sequence. In <6 hours, not months.
The Problem
An ML scientist and tech entrepreneur dedicated three months and a university research lab to designing a personalized cancer vaccine for his dog. NBC News called it a medical breakthrough. And it was.
But it required being an AI veteran with rare access to university infrastructure. That's not available to most dog owners and their vets. It was a one-off. We're making it the standard.
Design time
3 months
<6 hours
Accessibility
Research lab required
Any oncologist with sequencing data
Compute cost
$10,000+
~$15
How It Works
Standard variant call format from whole-genome or targeted sequencing. Paired tumor/normal or tumor-only.
NetMHCpan evaluates every peptide-allele combination across all mutations and lengths. Thousands of candidates per case, ranked by binding affinity and immunogenicity.
An AI model writes a veterinary clinical report. The pipeline outputs a codon-optimized mRNA FASTA ready for synthesis.
Output
Top 20 epitopes by composite score: IC50 binding affinity, immunogenicity prediction, variant allele frequency.
Veterinary-grade report with embedded charts, candidate reasoning, and recommended next steps.
IC50 bar chart and mutation landscape pie chart — publication-quality visuals for every case.
Codon-optimized construct with 5′UTR, Kozak, AAY-linked epitopes, 3′UTR, and poly-A(60) tail.
Follow a real canine mammary tumor case from somatic VCF to ranked neoantigen candidates, AI clinical report, and a synthesis-ready mRNA construct.
View Sample Case→