# Excelra data will support Boltz drug-discovery models

> Boltz will use curated GOSTAR measurements to train models that screen drug candidates for properties such as solubility and permeability.

Oossa · 2026-11-11 · https://oossa.com/en/excelra-data-will-support-boltz-drug-discovery-models

Boltz PBC has partnered with Excelra to use data from Excelra’s GOSTAR database in training ADMET models for its BoltzMol-1 drug-discovery system. The models estimate how compounds behave in the body and are intended to help filter out poor candidates before researchers buy or synthesize them.

Boltz says the GOSTAR-trained models matched or beat a leading open-source ADMET package on unfamiliar chemistry, with the biggest gains in solubility and permeability. The companies did not name the package or provide numerical scores in the announcement.

## The facts

- GOSTAR contains curated data on more than 11 million small molecules, Excelra says.
- Boltz says its prospective tests found confirmed binders or active compounds for six of 10 challenging targets, using 28–96 compounds per target.

## Why it matters

Drug researchers may be able to screen out some unsuitable compounds earlier, but the announcement does not establish how much the models improve real-world discovery or development outcomes.

## Sources & references

1. [Excelras GOSTAR™ unterstützt KI-Modelle von Boltz für die Wirkstoffforschung](https://www.prnewswire.com/news-releases/excelras-gostar-unterstutzt-ki-modelle-von-boltz-fur-die-wirkstoffforschung-302935987.html) – PR Newswire RSS, 2026-11-11

Last updated: 2026-11-11