Biocomputation · Artificial Intelligence · Data Science

Energy turned into knowledge

AI, biocomputation and data science applied to mRNA: variant classification and therapeutic design from its thermodynamic form factor — the physical axis no predictor uses. Sensitive data never leaves your infrastructure.

Λ = S ⊗ Φ
unified framework · signal ⊗ form
4publications · doi
42validated genes
0.99peak auc
1oepm patent
Tools · Herramientas

Open, glass-box and free

No signup. Each research line distills an interpretable physical observable into a tool that runs entirely on your machine — no GPU, no data leaving it.

Synonymous variants

ef-synonymous

The thermodynamic signal σ + G>A that matches the deep network. Available as a web tool, CLI and VEP plugin.

pip install ef-synonymous
Repeat expansion

energorna

Predicts onset and expansion risk in DM1, Huntington, SCA1 and Fragile X — beyond counting repeats, capturing interruption type and position. Validated across 4 diseases.

pip install energorna
01

Data Science

from raw data to an actionable result

02

Biocomputation

algorithms that speak molecular biology

03

Artificial Intelligence

cnn · llm · governed multi-agent

04

Thermodynamics

the mRNA form factor others don't see

Research line

Bio-AI — thermodynamic intelligence

The lab's first vertical. A biophysical platform that operates where REVEL and AlphaMissense are blind: the thermodynamic profile of the mRNA and, in particular, synonymous variants — a niche with no competitor. A shared engine, nearest-neighbor ΔG, feeds three fronts: variant classification, mRNA design and repeat-expansion disorders.

glass-box

The core finding: the signal reduces to an interpretable 2-variable physical model that matches the deep network — auditable as ACMG evidence.

Publications · bio-ai

Four publications with permanent DOI

Reviewable scientific evidence: three preprints and a technical benchmark report. A trilogy in progress, with peer review underway.

P-L1

EnergyFingerprint: Thermodynamic stacking profiles for missense variant classification

Multichannel 1D CNN, 42 validated genes, AUC 0.87–0.99. Zero-shot transferability gradient.

10.5281/zenodo.19831154
P-L2

mRNA Stacking Optimization: Codon design guided by nearest-neighbor thermodynamics

ΔG as an axis orthogonal to MFE (R²=0.144). A dual optimizer that outperforms commercial vaccines.

10.5281/zenodo.20228980
P1

Synonymous variants alter mRNA thermodynamic profiles: a CNN-based pathogenicity signal

AUC 0.683. Λ = S ⊗ Φ framework. The CNN reduces to 2 physical observables (signed ΔΔG + G>A bias), auditable as ACMG evidence (PP3/BP4).

10.5281/zenodo.20275792
TECH-2

Benchmark: EnergyFingerprint vs CADD, REVEL, AlphaMissense

Formal comparison across 7 genes. EF beats CADD in SCN1A (0.913 vs 0.832). Additive value demonstrated.

10.5281/zenodo.21309529
Contact

Jose Antonio Vilar Sánchez

QMetrika Labs
orcid: 0009-0008-1057-4223
github: @josevilar-qbioai
email: qmetrika[at]proton.me
write → github

Open science

Preprints with DOI, open-source code and two PyPI packages. Glass-box, no signup.

try ef-synonymous → try energorna →