Learning Protein Folding Energy Functions

An in-silico receptor-pharmacophore based multistep molecular docking and simulation study to evaluate the inhibitory potentials against NS1 of DENV-2

DENV-2 strain is the most fatal and infectious of the five dengue virus serotypes. The non-structural protein NS1 encoded by its genome is the most significant protein required for viral pathogenesis and replication inside the host body. Thus, targeting the NS1 protein and designing an inhibitor to limit its stability and secretion is a propitious attempt in our fight against dengue. Four novel inhibitors are designed to target the conserved cysteine residues (C55, C313, C316, and C329) and glycosylation sites (N130 and N207) of the NS1 protein in an attempt to halt the spread of the dengue infection in the host body altogether. Numerous computer-aided drug designing techniques including molecular docking, molecular dynamics simulation, virtual screening, principal component analysis, and dynamic cross-correlation matrix were employed to determine the structural and functional activity of the NS1-inhibitor complexes. From our analysis, it was evident that the extent of structural and atomic level fluctuations of the ligand-bound protein exhibited a declining trend in contrast to unbound protein which was prominently noticeable through the RMSD, RMSF, Rg, and SASA graphs. The ADMET analysis of the four ligands revealed a promising pharmacokinetics and pharmacodynamic profile, along with good bioavailability and toxicity properties. The proposed drugs when bound to the targeted cavities resulted in stable conformations in comparison to their unbound state, implying they have good affinity promising effective drug action. Thus, they can be tested in vitro and used as potential anti-dengue drugs.Communicated by Ramaswamy H. Sarma.

Elucidation of the structural dynamics of mutations in PHB2 protein associated with growth suppression and cancer progression

Prohibitin is a multifunctional protein that plays an important role in numerous cellular processes. Membrane-associated mitochondrial prohibitin complex is made up of two subunits, PHB1 and PHB2 which are ubiquitously expressed and analogous to each other. High levels of prohibitin expression have consequently been found in esophageal cancer, endometrial adenocarcinoma, gastric cancer, hepatocellular carcinoma, breast cancer and bladder cancer. The aim of this study is to analyse two-point mutation PHB2_MT1(I → A) and PHB2_MT2(I → P), their effect on PHB2 protein and its effect on formation of mitochondrial complex. It is a residual level study, based on current experimental validation. To establish the effects of the two-point mutations, computational approaches such as molecular modelling, molecular docking, normal mode simulation, molecular dynamics simulations and MM/GBSA were used. An analysis of the energy dynamics of both unbound and complex proteins was conducted to elucidate how mutations impact the energy distribution of PHB2. Our study confirmed that the two mutations decreased the overall stability of PHB2. This was evidenced by heightened atomic fluctuations within the mutated region, accompanied by elevated deviations observed in RMSD and Rg values. Furthermore, these mutations were correlated with a decline in the organization of secondary structural elements. The mutations in PHB2_MT1 and PHB2_MT2 resulted in formation a less stable prohibitin complex. Thus, PHB1 and PHB2 may act as molecular target or novel biomarkers for therapeutic intervention in numerous forms of malignancies.

Learning Organizations of Protein Energy Landscapes: An Application on Decoy Selection in Template-Free Protein Structure Prediction

The protein energy landscape, which lifts the protein structure space by associating energies with structures, has been useful in improving our understanding of the relationship between structure, dynamics, and function. Currently, however, it is challenging to automatically extract and utilize the underlying organization of an energy landscape to the link structural states it houses to biological activity. In this chapter, we first report on two computational approaches that extract such an organization, one that ignores energies and operates directly in the structure space and another that operates on the energy landscape associated with the structure space. We then describe two complementary approaches, one based on unsupervised learning and another based on supervised learning. Both approaches utilize the extracted organization to address the problem of decoy selection in template-free protein structure prediction. The presented results make the case that learning organizations of protein energy landscapes advances our ability to link structures to biological activity.