Our deep learning solutions are ideal for application to drug discovery, where the available data are particularly sparse and noisy. Alchemite™ can learn simultaneously across all experimental endpoints to target high-quality compounds and prioritise experimental resources. Our technology can be applied to projects in both small molecule and biologics.
In collaboration with Optibrium, our partners in small molecule drug discovery, Alchemite™ has been demonstrated to make more effective use of limited data than conventional QSAR, and confidently highlight the most accurate predictions on which to base your decisions. For more information on small molecule drug discovery applications, please visit Cerella at Optibrium.com.
Alchemite™ for Antimalarial Drug Discovery
Challenge: To identify a new antimalarial compound against a novel antimalarial mechanism of action. Available data covers multiple activity assays, but only a small proportion have been measured in practice.
Solution: Alchemite™ outperforms conventional QSAR and other AI models and can accurately predict bioactivity from sparse data (<10% complete).
Outcome: Alchemite™ confidently predicted a compound generated automatically by Optibrium’s StarDrop™ software. Compound synthesised and tested by the Open Source Malaria group and was the only entry that demonstrated potency against the target.
For more information on small-molecule drug discovery using Alchemite™, visit https://www.optibrium.com
Read Press Release here.
Related resources & news
Swipe to navigate articles