EQUIBIND: A geometric deep learning-based protein-ligand binding prediction method

  • Li Yuze
    Department of Medical Chemistry, School of Pharmacy, Qingdao University, Qingdao, Shandong, China.
  • Li Li
    Department of Medical Chemistry, School of Pharmacy, Qingdao University, Qingdao, Shandong, China.
  • Wang Shuang
    Department of Medical Chemistry, School of Pharmacy, Qingdao University, Qingdao, Shandong, China. Department of Stomatology, Huangdao District Central Hospital, Qingdao, Shandong, China.
  • Tang Xiaowen
    Department of Medical Chemistry, School of Pharmacy, Qingdao University, Qingdao, Shandong, China.

抄録

<p>Structure-based virtual screening plays a critical role in drug discovery. However, numerous docking programs, such as AutoDock Vina and Glide, are time-consuming due to the necessity of generating numerous molecular conformations and executing steps like scoring, ranking, and refinement for the ligand-receptor complexes. Consequently, achieving rapid and reliable virtual screening remains a noteworthy challenge. Recently, a team of researchers from Massachusetts Institute of Technology, led by Stärk et al., developed an SE(3)-equivariant geometric deep learning based protein-ligand binding prediction approach, EQUIBIND. In comparison to conventional docking methods, EQUIBIND has the capacity to predict the binding modes of small molecules with target proteins rapidly and precisely. It presents an innovative resolution for high-throughput screening of drug-like compounds. </p>

収録刊行物

  • Drug Discoveries & Therapeutics

    Drug Discoveries & Therapeutics 17 (5), 363-364, 2023-10-31

    特定非営利活動法人 バイオ&ソーシャル・サイエンス推進国際研究交流会

参考文献 (7)*注記

もっと見る

詳細情報 詳細情報について

問題の指摘

ページトップへ