Automatic drug design method and system, computing equipment and computer readable storage medium

Abstract

The invention discloses an automatic medicine design method and system, computing equipment and a computer readable storage medium. The method comprises the following steps of decomposing a target lead compound into fragments with synthesizable modules, and sequentially inputting the fragments into a trained drug design machine learning model for sampling, and reassembling the new fragment outputby the drug design machine learning model to obtain a new lead compound. According to the automatic drug design provided by the invention, the performance in the aspects of molecule generation effectiveness and uniqueness is greatly improved, and molecules with high novelty, strong synthesizability and strong druggability can be generated; molecules can be easily generated in a high molecular weight region; the method can be repeatedly used for different target point lead compound generation scenes only by using a specific data set to train once; the local structure of the immobilized compoundcan be easily realized, and the rest parts are optimized.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to an automatic drug design method, system, computing equipment and computer-readable storage medium. Background technique [0002] Designing lead compounds with desirable properties is a central task in the drug discovery phase. In fast-follow and similar (Me-too) drug design scenarios, the traditional process requires the collection of a large number of papers and patents, and on the basis of reading and understanding by medicinal chemists, design novel, synthesizable, and druggable drugs Strong compounds, and validated by chemical synthesis and biological characterization. [0003] Molecular Generation is an automatic drug design method based on deep generative learning that has developed rapidly in recent years. By letting the model learn the SMILES (a compound structure represented by a string) or Molecular Graph (atoms and chemical bond connections represented by m...

AI Technical Summary

Problems solved by technology

[0005] Problem 1, the validity and uniqueness of molecules generated by the L2L framework model are low

[0006] Problem 2, the high novelty of molecules generated by the L2L framework model is low, and it is difficult to generate compounds with normal structures in the chemical space far away from the target molecule (molecular similarity Tc<0.4)

[0007] Problem 3, the L2L framework model is difficult to generate structurally normal compounds with a molecular weight (Molecular Weight, MW) greater than 500, so it cannot be applied to some lead compound design scenarios that require high molecular weight

[0008] Question 4. The L2L framework model cannot be reused across targets. When applied to different targets, it is necessary to add known active compounds of the target target for training to generate new structural compounds, resulting in waste of time, manpower, and material resources.

[0009] Problem 5, the L2L framework cannot allow the model to fix a certain part of the compound structure, so that automatic sampling can be performed in other parts

[0010] The above problems limit the practical application value of molecular generative models using L2L framework

Images