ACFIS: a web server for fragment-based drug discovery

Abstract

In order to foster innovation and improve the effectiveness of drug discovery, there is a considerable interest in exploring unknown 'chemical space' to identify new bioactive compounds with novel and diverse scaffolds. Hence, fragment-based drug discovery (FBDD) was developed rapidly due to its advanced expansive search for 'chemical space', which can lead to a higher hit rate and ligand efficiency (LE). However, computational screening of fragments is always hampered by the promiscuous binding model. In this study, we developed a new web server Auto Core Fragment in silico Screening (ACFIS). It includes three computational modules, PARA_GEN, CORE_GEN and CAND_GEN. ACFIS can generate core fragment structure from the active molecule using fragment deconstruction analysis and perform in silico screening by growing fragments to the junction of core fragment structure. An integrated energy calculation rapidly identifies which fragments fit the binding site of a protein. We constructed a simple interface to enable users to view top-ranking molecules in 2D and the binding mode in 3D for further experimental exploration. This makes the ACFIS a highly valuable tool for drug discovery. The ACFIS web server is free and open to all users at http://chemyang.ccnu.edu.cn/ccb/server/ACFIS/.

Figure 1.

The interplay of the three ACFIS modules. PARA_GEN runs in a serial mode, whereas CORE_GEN and CAND_GEN run in parallel mode.

Figure 2.

Workflow of ACFIS. The user input is shown in blue. The arrows denote the computational process.

Figure 3.

A case study of cytochrome bc₁ complex inhibitors design with ACFIS. Input is the inhibitor WF3 presented in the crystallized structure of cytochrome bc₁ complex (PDB code: 3TGU). The composite computation includes parameter generation, fragment deconstruction, core fragment identification, fragment linking, structure optimization, molecular dynamic simulation and binding free energy calculation, which can be processed automatically by this web server. Outputs are the top five hit molecules.