LIA
Genomics
Accelerating Genetic Variant Classification
Origin
“LIA is named after my daughter, diagnosed with SYNGAP1 — a rare genetic condition affecting brain development. The evidence existed. The tools to assemble it quickly did not.”
When a genetic variant is found, clinicians must determine whether it causes disease — a process that currently takes hours.
Who it's for
01.
Clinical Geneticists
From hours of manual curation to a fully evidenced, ACMG-classified interpretation in seconds, ready for clinician sign-off.
02.
Diagnostic Laboratories
Reproducible, provenance-tagged annotation at VCF scale. Every report traceable, every source versioned.
03.
Rare Disease Researchers
Automatic evidence assembly across ClinVar, gnomAD, and PubMed. Spend time on discovery, not chasing evidence.
CapabilitiesFour steps from notation to clinical-grade report
01.
Resolve
Any notation — HGVS, rsID, coordinates, gene symbol. Build and transcript ambiguity surfaces as warnings, never hidden.
02.
Annotate
Fast Streaming pipeline. ClinVar, gnomAD LOEUF/pLI, AlphaMissense. Full provenance on every field. 60–150× faster than current tools.
03.
Review
AI-grounded literature search over PubMed and internal knowledge bases. Every result is backed by a specific retrieved record. Uncited claims are excluded by design.
04.
Classify
Rule-based ACMG/AMP 2015 classifier. Every applied criterion carries a cited rationale. Non-assessable criteria are stated explicitly, never silently omitted. Reproducible clinical-grade report included.
Source available at launch · Self-hostable · Patient data stays yours
Performance
2–8 h
30s0s
per variant
10–26 h
15 min0 min
10M variant VCF
1–3 h
5 min0 min
literature review per VUS
Stay in the Loop
Get product updates, launch announcements, and technical notes delivered to your inbox.
Join researchers and geneticists shaping the future of variant classification.