A method and device for virtual screening of small molecule drugs based on unsupervised domain adaptation

Abstract

The invention provides a small molecule drug virtual screening method based on unsupervised domain adaptation, which predicts new drug target information according to known drug target information, and performs virtual screening; the method includes constructing target domain and source domain drug target data sets Generative models; build and train unsupervised domain-adapted drug virtual screening models; build small-molecule drug virtual screening models targeting target-domain drug targets. In the process of model construction, it can effectively solve the problem that the virtual screening model is not easy to build or the construction is not ideal due to insufficient number of samples and information. Better virtual screening is of great help in the research of new drug targets and lead compounds for rare diseases. The purpose of improving the efficiency of screening and saving time and huge capital expenditure can be achieved. When applied to the field of drug screening, it can also meet the requirements of biological activity prediction of drug target ligands.

Description

technical field [0001] The invention relates to computer-aided drug design, in particular to a method and device for virtual screening of small molecule drugs based on unsupervised domain adaptation. Background technique [0002] In recent years, the development of new drugs for new drug targets or rare diseases has become a research hotspot in recent years. The development of new drugs needs to use high-throughput experimental technology to determine the biological activity value of the target and compound action in a huge database of compounds, and then screen the lead compounds. However, experimental methods are time-consuming and labor-intensive, and worse, the number of compounds available is often very limited, and not all drug targets are suitable for high-throughput screening experiments. Therefore, computational-based virtual screening, which simulates the interaction between target targets and drug candidates, has been widely used in drug design, of which machine ...

AI Technical Summary

Problems solved by technology

However, the virtual screening of ligands for new drug targets or rare diseases sometimes has insufficient or even missing training samples, and good models are often not obtained when performing virtual screening on them

Images