Skip to main content
Fig. 6 | Journal of Cheminformatics

Fig. 6

From: Machine intelligence-driven framework for optimized hit selection in virtual screening

Fig. 6

Binding pocket investigation and quantification for the target protein. Binding pocket assessment and crucial amino acid residues designation is vital for pharmacological activity by liable molecules. Herein, we displayed the binding pocket investigation and quantification layout. We first retrieved the 3D structure of CXCR4 (a) protein from the PDB database and bound ligand (PDB ID: 3ODU). The CXCR4 structure was subjected to the Cavity program to assess vacant ligand-binding pocket (red) and grid points (violet) as represented in b and c. For binding residue quantification, the results of the Cavity program were used as input by Pocket v3 that result in the amino acid residues location (d) along with probable types of interaction. The blue colored spheres symbolize hydrogen bond donor, red are hydrogen bond acceptor, and pink represents hydrophobic interaction residues. In summary, the active residues comprise the active binding pocket for CXCR4 protein (e) along with residue location. The information generated will serve as evaluation for assessment of PS-driven DNNs framework of A-HIOT with other machine learning algorithms

Back to article page