Glass-box thermodynamic scoring of synonymous variants (signed stacking ΔΔG + G>A)
Project description
ef-synonymous
Scoring glass-box de patogenicidad de variantes sinónimas a partir de un único observable físico interpretable: el ΔΔG de apilamiento nearest-neighbor con signo (parámetros Turner de RNA) más el sesgo mutacional G>A.
Un modelo de 2 variables que reproduce a la CNN dedicada del Paper 1
(AUC 0.671 pooled / 0.680 leave-genes-out ≈ CNN 0.683). Inferencia solo con
numpy — sin PyTorch, sin GPU, sin llamadas a servicios externos. Es la misma
lógica del prototipo web y de core/synonymous_sigma_predictor.py, verificada a
1e-6.
⚠ Research use only. No es un dispositivo médico. Los umbrales de mapeo ACMG son ilustrativos y no están calibrados clínicamente.
Instalación
pip install -e ef-synonymous-tool # desde el repo
# o, publicado: pip install ef-synonymous
CLI
# CDS literal + HGVS
ef-syn score --cds ATGGTC...TAG --hgvs c.213G>A --gene CACNA1C
# desde un FASTA local
ef-syn score --cds-file NM_000719.cds.fasta --pos 213 --ref G --alt A
# descargando el CDS de un transcrito (Ensembl REST, red del usuario)
ef-syn score --transcript ENST00000399655 --hgvs c.213G>A
# dossier auditable JSON (base del argumento SaMD)
ef-syn score --cds-file cds.fasta --hgvs c.213G>A --json
Salida humana:
EF-Synonymous · c.213G>A (CACNA1C)
score_sigma (P patogénica): 0.688
ACMG: PP3 (supporting pathogenic) [umbrales ilustrativos, sin calibrar]
codón: GCG(A) -> GCA(A) — sinónima ✓
glass-box (contribuciones al logit):
intercepto -0.062
sigma_signed 1.67 +0.071
is_GtoA 1 +0.782
logit = 0.791 -> P = 0.688
Librería
from ef_synonymous import SynonymousSigmaPredictor
clf = SynonymousSigmaPredictor.load()
clf.predict_proba(cds, cds_pos=213, ref="G", alt="A") # 0.688...
clf.score(cds, 213, "G", "A", gene="CACNA1C") # dossier completo (dict)
Plugin Ensembl VEP
cp vep_plugin/EFSynonymous.pm ~/.vep/Plugins/
vep -i input.vcf --plugin EFSynonymous
Añade EF_sigma, EF_score y EF_acmg a las variantes synonymous_variant.
El plugin delega en la CLI (mismo motor). Para escala, la vía de producción es
una cache precomputada por transcrito o un endpoint REST — este plugin es la
implementación de referencia de la integración.
Verificación
python -m pytest tests/ # o: python tests/test_ground_truth.py
Reproduce el valor validado (CACNA1C c.213G>A → σ=1.67, P=0.688068), el mismo que verifica la implementación JavaScript del prototipo web.
Método y licencia
Código bajo licencia MIT. El método de huella termodinámica del mRNA está cubierto por la patente P202630522 (OEPM); su uso comercial requiere licencia. Paper 1: Zenodo 10.5281/zenodo.20275792.
QMetrika Labs · Jose Antonio Vilar Sánchez.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file ef_synonymous-0.2.0.tar.gz.
File metadata
- Download URL: ef_synonymous-0.2.0.tar.gz
- Upload date:
- Size: 12.0 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b22746fe933f4cdd98fc27d49291c5ecae7bdf0a24b6eb745edfb089402ef144
|
|
| MD5 |
ba9308130611dfb21f40586d274db4da
|
|
| BLAKE2b-256 |
8c4956fb64d1b7637c3fbf9708338f200fcb3077f92e3f2a9115989e41c0a44a
|
Provenance
The following attestation bundles were made for ef_synonymous-0.2.0.tar.gz:
Publisher:
publish.yml on josevilar-qbioai/ef-synonymous
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
ef_synonymous-0.2.0.tar.gz -
Subject digest:
b22746fe933f4cdd98fc27d49291c5ecae7bdf0a24b6eb745edfb089402ef144 - Sigstore transparency entry: 2194743064
- Sigstore integration time:
-
Permalink:
josevilar-qbioai/ef-synonymous@41524664235f65930312035222c83ccd5d69fe27 -
Branch / Tag:
refs/tags/v0.2.0 - Owner: https://github.com/josevilar-qbioai
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@41524664235f65930312035222c83ccd5d69fe27 -
Trigger Event:
push
-
Statement type:
File details
Details for the file ef_synonymous-0.2.0-py3-none-any.whl.
File metadata
- Download URL: ef_synonymous-0.2.0-py3-none-any.whl
- Upload date:
- Size: 12.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
daed0ff86a45a70007aba07e1a23ee492052f6c35c36bf116b79fd07003000ce
|
|
| MD5 |
135c2d5bea8b11f9068c4ccd41d38a66
|
|
| BLAKE2b-256 |
e943e22b3a229f6bb2f765b734e431093a0618053af1515de126a15e8225614e
|
Provenance
The following attestation bundles were made for ef_synonymous-0.2.0-py3-none-any.whl:
Publisher:
publish.yml on josevilar-qbioai/ef-synonymous
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
ef_synonymous-0.2.0-py3-none-any.whl -
Subject digest:
daed0ff86a45a70007aba07e1a23ee492052f6c35c36bf116b79fd07003000ce - Sigstore transparency entry: 2194743080
- Sigstore integration time:
-
Permalink:
josevilar-qbioai/ef-synonymous@41524664235f65930312035222c83ccd5d69fe27 -
Branch / Tag:
refs/tags/v0.2.0 - Owner: https://github.com/josevilar-qbioai
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@41524664235f65930312035222c83ccd5d69fe27 -
Trigger Event:
push
-
Statement type: