Eurofins Discovery, part of the Eurofins network of laboratories and an industry-leading provider of products and services for drug discovery research, announces the launch of DiscoveryAI SAFIRE (Suite of ADMET Predictions For In Silico Refinement and Evaluation). SAFIRE is an advanced platform that leverages proprietary datasets, artificial intelligence (AI), and machine learning (ML), offering a unique capability for expediting discovery. SAFIRE is now available for customer use through the Eurofins Discovery website.
Eurofins Discovery’s experienced informatics experts collected vast, high-quality and diverse datasets that provide a unique approach in predicting the ADMET (Absorption, Distribution, Metabolism, Excretion and Toxicity) properties of molecules. It sets a new standard in predictive analytics.
SAFIRE utilises data from Eurofins Discovery’s preeminent BioPrint database, curated over years of meticulous research and data accumulation. The result: a more powerful research tool surpassing the limits that exist in other predictive systems currently in the market, as demonstrated in a recent publication in the Future Medical Chemistry journal.
Also Read: Trinity Life Sciences Launches Brand Insights AI
Designed with end-users in mind, SAFIRE provides essential features such as scoring functionality to aid in decision-making and a comprehensive suite of predictive ADME models. It’s proven effective in accelerating drug discovery projects from initial hit discovery through to lead optimisation.
How does SAFIRE go beyond existing predictive system limits in the market? Published results of AI advances demonstrated how Eurofins Discovery’s models can be used to predict ADMET properties of molecules quickly and efficiently, when trained with BioPrint database proprietary data along with public datasets utilised by the SAFIRE platform. This approach has proven improved model performance as well as broadened the chemical space covered by the models, achieving Matthews Correlation Coefficient values greater than 0.4, and accuracies above 80%.
SOURCE: BusinessWire
