Dioclea Labs develops AI-powered sampling algorithms that improve the accuracy of predicting experimental results in drug discovery. Their technology enhances predictions of binding affinity and efficiency, accelerating the identification of promising drug candidates.
Funding
Funding not disclosed
Founders
Product
Problem
Traditional drug discovery methods often struggle to accurately predict experimental results, leading to inefficient identification of promising drug candidates. Accurately estimating free energy surfaces, which are crucial for understanding protein behavior and ligand binding, can be computationally expensive and time-consuming. This gap between simulations and real-world experimental outcomes hinders rational drug design.
Solution
Dioclea Labs offers AI-powered sampling algorithms that enhance the accuracy and speed of predicting experimental results in drug discovery. Their technology leverages evolutionary and structural information, combined with advanced machine learning, to improve predictions of binding affinity, efficiency, and other functionally coupled observables. By estimating free energy surfaces of binding or conformational changes in a fraction of the time, Dioclea Labs enables comprehensive studies without limiting scope. The company's approach retains atomistic detail necessary for rational drug design while accurately predicting large-scale experimental observables, bridging the gap between simulations and reality.
Target Audience
Dioclea Labs primarily serves pharmaceutical companies and research institutions involved in drug discovery and development.
Features
- AI-fueled sampling algorithms for estimating free energy surfaces of binding and conformational change
- Integration of evolutionary and structural information to improve prediction accuracy
- Iterative exploration algorithms seeded by coevolutionary models
- Identification and mapping of functional states by leveraging evolutionary information
- AI generation of novel compounds by analyzing complementarity requirements in the drug binding site
- Cross-reactivity experiments and in-silico screening of compounds across different targets
- Molecular dynamics simulations accelerated using enhanced sampling algorithms and on-the-fly neural network evaluation
- Dimensionality reduction using information-greedy algorithms