AUTHOR=Sun Dingkang , Xu Lulu , Tong Mengfan , Wei Zhao , Zhang Weitong , Liang Jialong , Liu Xueying , Wang Yuwei 

TITLE=De novo design of mIDH1 inhibitors by integrating deep learning and molecular modeling

JOURNAL=Frontiers in Pharmacology

VOLUME=Volume 15 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2024.1491699

DOI=10.3389/fphar.2024.1491699

ISSN=1663-9812

ABSTRACT=Mutations in the IDH1 gene have been shown to be an important driver in the development of acute myeloid leukemia (AML), gliomas and certain solid tumors, which are a promising target for cancer therapy. The aim of this study was to generate candidate compounds with potential activity and good drug-likeness using the bidirectional recurrent neural network (BRNN) model and scaffold hopping, followed by virtual screening and molecular dynamics simulations. The BRNN model ultimately generated 3890 new compounds, while scaffold hopping finally generated 3680 new compounds. The molecules generated by both approaches were evaluated by PCA, QED, SA analysis and molecular docking, which was found that the molecules generated by the BRNN model had superior molecular diversity, druggability, synthesizability and docking scores. Therefore, 3890 new compounds generated by BRNN model were screened using glide-based virtual screening. Ultimately, 10 structurally diverse compounds were retained, all of which showed the potential to become candidate drugs in terms of ADME properties. Molecular dynamics simulations of 6 small molecules with better scores than positive compounds showed that the RMSD of the four systems of M1 and M2, M3 and M6 remained stable, and had local flexibility and compactness similar to the positive drugs. Finally, the free energy decomposition results showed that compound M1 exhibited the best binding properties in all energy aspects and was the best candidate molecule among the 6 compounds. This study is the first attempt to use deep learning to design mIDH1 inhibitors, which provides theoretical guidance for the design of mIDH1 inhibitors.