Excelra's GOSTAR™ Enhances Boltz AI Models for Drug Discovery Efficiency

In an innovative move within the pharmaceutical sector, Boltz PBC has teamed up with Excelra to enhance its artificial intelligence models aimed at drug discovery. This partnership sees the utilization of Excelra's GOSTAR™ data, which plays a crucial role in providing curated measurements that significantly improve the predictive capabilities of Boltz's AI models.

The collaboration focuses on the integration of curated kinetic solubility, lipophilicity, and Caco-2 permeability data from GOSTAR™ into Boltz's ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) models known as BoltzMol-1. This initial filtering mechanism is designed to sift out less favorable compounds before they enter the more resource-intensive phases of procurement or synthesis.

Notably, results from Boltz’s recent technical report reveal that the AI models, trained with GOSTAR™ data, performed exceptionally well. They were able to confirm active compounds for six out of ten challenging target structures using only 28 to 96 compounds per target. This approach starkly contrasts with traditional high-throughput screenings that often require thousands to millions of compounds but yield success rates below 0.15%.

Gabriele Corso, the co-founder and CEO of Boltz PBC, expressed the significance of GOSTAR™ in their collaborative efforts: "The scope of the GOSTAR™ dataset allows us to build models capable of generalizing across chemical processes far beyond typical training data. It has become an essential input for our modeling at Boltz, enabling us to save considerable time in our research."

Excelra’s GOSTAR™ database provides comprehensive insights into over 11 million small molecules and their biological attributes. This extensive collection, which is ISO 9001 certified, has been engineered specifically to aid organizations in the life sciences through advanced data science and AI analytics.

Anandbir Brar, Managing Director of Excelra, also highlighted the pivotal role of GOSTAR™ in the development of AI-supported drug discovery: "Providing solubility and kinetic data at this scale is a gap that public sources often fail to fill. GOSTAR™ was created to support prospective model validation rather than simply rely on retrospective benchmarks."

This validated data-driven approach is poised to transform how organizations conduct drug discovery, promoting efficiency and accuracy in the identification of viable compounds. As Boltz continues to refine its AI systems, the partnership with Excelra signals a progressive advancement in the realm of drug research and development.

Excelra has established itself as a key partner for data science and AI across leading life science organizations, further enhancing the pace and intelligence of drug research and development efforts. As this partnership unfolds, the implications for faster and more effective drug discovery could be profound, contributing to innovations in therapeutic solutions.

For additional information about Excelra and the GOSTAR™ database, you can visit excelra.com/databases/gostar. Boltz PBC, on the other hand, epitomizes applied AI research with a mission to democratize access to cutting-edge technologies in biological discoveries. More details about their initiatives can be accessed at boltz.bio.