Skip to main content

Omnix

Omnix runs predictive machine learning, including antibiotic-resistance prediction grounded in established resistance-gene databases. Every prediction is explainable, and every output is marked research-only.

ResearchersClinical researchPublic healthPharma enterprise
LumenPraxisVeraxaNexoraNovoraOrisOmnixstep 7SentinelPrognosLabtrack

Inputs

  • Analytical results or genomic data from Novora
  • Prediction task (including AMR genotype-to-phenotype)
  • Reference database selection

Outputs

  • Predictions with mandatory SHAP feature attributions
  • AMR resistance calls referenced to source databases
  • IntendedUse.RESEARCH_ONLY annotation on every output

How the AI suggests, and where you decide

Omnix predicts and explains; it never auto-acts on a prediction. SHAP attributions accompany every output so the researcher can see why a call was made before deciding what to do with it.

Illustrative

Omnix in practice

Ciprofloxacin resistance trajectory, E. coli
Illustrative forecast of E. coli resistance to ciprofloxacin rising from 20 percent in 2018 to 35 percent in 2024, projected to about 41 percent by 2027 with a widening confidence interval.10%20%30%40%50%201820202022202420262027forecast →
Observed Projected Confidence interval

Citation grounding

AMR predictions reference CARD, ResFinder, AMRFinderPlus, PointFinder, and BV-BRC. SHAP interpretability is mandatory, not optional, on every prediction.

For the technical reader

Technical details

Each capability is one link in the closed loop. See the whole chain end to end.