Computer implemented method for calculating a bound-like structure of a protein
The computer-implemented method predicts protein bound-like structures by identifying quasi-rigid domains and incorporating charged residues, improving prediction accuracy and speed, thus reducing drug design timelines.
Patent Information
- Application Number
- PCT/IB2024/056224
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-06-26
- Filing Date
- 2024-06-26
- Publication Date
- 2025-01-02
AI Technical Summary
Current methods for predicting the bound-like structure of proteins rely on in vitro experiments, which are time-consuming and lack precision, especially when dealing with proteins having multiple quasi-rigid domains and charged residues.
A computer-implemented method that identifies quasi-rigid domains and includes electrically charged and uncharged residues in the binding site to enhance the prediction of bound-like conformations, using inertia planes, collective variables, and bias-exchange well-tempered metadynamics simulations to improve the accuracy and speed of docking calculations.
This approach significantly reduces the lead time for drug design by providing a reasonable number of accurate putative bound-like conformations, enhancing the precision of protein structure predictions and facilitating faster drug development.
Smart Images
Figure IB2024056224_02012025_PF_FP_ABST