Detection and characterization of protein cavities using parallel computing and molecular descriptors

• Grant number: 18/00629-0
• Support Opportunities: Regular Research Grants
• Duration: October 01, 2018 - March 31, 2021
• Field of knowledge: Interdisciplinary Subjects
• Principal Investigator: Paulo Sergio Lopes de Oliveira
• Grantee: Paulo Sergio Lopes de Oliveira
• Host Institution: Centro Nacional de Pesquisa em Energia e Materiais (CNPEM). Ministério da Ciência, Tecnologia e Inovação (Brasil). Campinas , SP, Brazil
• Associated researchers: João Victor da Silva Guerra ; José Geraldo de Carvalho Pereira ; Rodrigo Vargas Honorato

Abstract

The prospection and characterization of binding sites are essential tasks in the structural analysis of proteins, allowing a better understanding of the complex macromolecular interactions. Topological and physicochemical characteristics presented at the contact interface dictate the functions of proteins. The advancement of proteomics in the last decade has brought to the surface a virtually inexhaustible repository of potential targets for drugs and the development of computational methods for prospecting and characterization of binding sites in proteins is essential for drug discovery and enhancement in addition to deepening the knowledge about the function of a particular protein. Programs developed for cavity prospecting are divided into three categories according to the applied method: geometric, energetic or evolutionary, as well as combinations of them. The KVFinder program, developed in our laboratory, uses a geometric approach, has free and open code and is distributed free of charge. In order to ensure maximum usability, a graphical interface was developed for the PyMOL visualization software. The prospection is done through voxels, applying a double probe system, resulting in the insertion of the protein to be analyzed in two distinct three-dimensional grids, which in the end are subtracted for the characterization of the cavities. The current implementation of KVFinder is based on serial computing and the geometric nature of the algorithm generates a high demand for computational resources. This research proposal aims to improve the KVFinder program with the following objectives: 1) Creation of a parallel version of the software, provided that the matrix operations of the algorithm can be applied in each voxel independently; 2) Use of physicochemical characteristics to identify relevant cavities and guide the rational design of drugs, Since the interactions of a cavity are are governed by these characteristics, which in the end, determine the biological function of the protein; 3) Creation of a routine for the evaluation of cavities in a trajectory obtained by molecular dynamics simulation. Since it is known that the dynamics of the cavities is an essential component for the understanding of their interactions, and the structural flexibility allows the adaptation of the proteins to the ligands, it makes the dynamic evaluation of the topological modifications of the cavities is extremely relevant, allied to the fact that the parallelization of the algorithm obtained in objective 1 makes this objective feasible and; 4) The implementation and distribution of KVFinder as a web platform and as a web service, allowing the analysis of large scale data for the user not specialized in bioinformatics. For example, prospecting and characterization of binding sites in various proteins of the same family or group, in order to find the conservation of these sites, would become plausible to community researchers with no programming experience. (AU)

Scientific publications

(6)

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DA SILVA GUERRA, JOAO VICTOR; RIBEIRO-FILHO, HELDER VERAS; JARA, GABRIEL ERNESTO; BORTOT, LEANDRO OLIVEIRA; DE CARVALHO PEREIRA, JOSE GERALDO; LOPES-DE-OLIVEIRA, PAULO SERGIO. pyKVFinder: an efficient and integrable Python package for biomolecular cavity detection and characterization in data science. BMC Bioinformatics, v. 22, n. 1, DEC 20 2021. (18/00629-0)

RIBEIRO-FILHO, V, HELDER; COIMBRA, LAIS D.; CASSAGO, ALEXANDRE; ROCHA, REBECA P. F.; DA SILVA GUERRA, JOAO VICTOR; DE FELICIO, RAFAEL; CARNIELI, CAROLINA MORETTO; LEME, LUIZA; PADILHA, ANTONIO CLAUDIO; PAES LEME, ADRIANA F.; et al. Cryo-EM structure of the mature and infective Mayaro virus at 4.4 angstrom resolution reveals features of arthritogenic alphaviruses. NATURE COMMUNICATIONS, v. 12, n. 1, MAY 24 2021. (18/00629-0, 18/03917-6, 17/15340-2)

ZEN PETISCO FIORE, ANA PAULA; RODRIGUES, ANA MARIA; RIBEIRO-FILHO, HELDER VERAS; MANUCCI, ANTONIO CARLOS; DE FREITAS RIBEIRO, PEDRO; SILVA BOTELHO, MAYARA CAROLINNE; VOGEL, CHRISTINE; LOPES-DE-OLIVEIRA, PAULO SERGIO; PAGANO, MICHELE; BRUNI-CARDOSO, ALEXANDRE. Extracellular matrix stiffness regulates degradation of MST2 via SCF beta TrCP. BIOCHIMICA ET BIOPHYSICA ACTA-GENERAL SUBJECTS, v. 1866, n. 12, p. 11-pg., 2022-09-01. (17/25437-3, 17/18641-3, 18/00629-0, 14/25832-1, 14/10492-0, 17/18067-5, 19/26767-2)

DA SILVA GUERRA, JOAO VICTOR; RIBEIRO FILHO, HELDER VERAS; BORTOT, LEANDRO OLIVEIRA; HONORATO, RODRIGO VARGAS; DE CARVALHO PEREIRA, JOSE GERALDO; LOPES-DE-OLIVEIRA, PAULO SERGIO. ParKVFinder: A thread-level parallel approach in biomolecular cavity detection. SOFTWAREX, v. 12, JUL-DEC 2020. (18/00629-0)

GUERRA, JOAO V. S.; RIBEIRO-FILHO, HELDER, V; PEREIRA, JOSE G. C.; LOPES-DE-OLIVEIRA, PAULO S.. KVFinder-web: a web-based application for detecting and characterizing biomolecular cavities. Nucleic Acids Research, v. 51, n. W1, p. 9-pg., 2023-05-04. (18/00629-0)

GUERRA, JOAO V. S.; ALVES, LUIZ F. G.; BOURISSOU, DIDIER; LOPES-DE-OLIVEIRA, PAULO S.; SZALOKI, GYORGY. Cavity Characterization in Supramolecular Cages. JOURNAL OF CHEMICAL INFORMATION AND MODELING, v. 63, n. 12, p. 14-pg., 2023-05-02. (18/00629-0)