With our powerful auto-feature-extraction algorithms, we can find more information from structures than traditional Fingerprints do, and the accuracy rate of our prediction is much better than other solutions.
Using our reliable prediction services at the early stage will let you know which compound may fail at the pre-clinical stage. You can then optimize it or give it up early to avoid unnecessary time costs.
Are you tired of black-box machine learning models? Our explainable AI tells you why the compounds will fail in the future pre-clinical tests. The AI works even better with our consultation on your final decisions.
We prepare the highly integrated AI services for you: our ADDD(AI-Driven Drug Development) platform can automatically optimize the compound with several endpoints simultaneously, speeding up the drug development. Our consultants guide you to integrate AI into your drug development and develop your proprietary machine learning models that bring you with the candidates.
VIRTUALMAN contains dozens of endpoints, such as ADME (e.g. Caco-2 permeability, CYP inhibition), toxicity(e.g. hERG, mutagenicity), pharmacokinetics(e.g. volume of distribution, clearance) etc., helping you find druggable compounds instantly.
Our prediction service is highly performed and widely applicable. However, if you have some in-vivo/in-vitro data, we can combine your data with ours to generate even more accurate prediction. Customized services will be charged based on the scope of your projects.
ADDD platform
VIRTUALMAN for ADMET/PK
Customized AI
