A machine learning approach for virtual screening of histone deacetylase inhibitor compounds using aromatic signatures

• Alessandra Gomes Cioletti Universidade Federal de Minas Gerais (UFMG) https://orcid.org/0009-0008-3486-3548
• Diego Mariano Universidade Federal de Minas Gerais (UFMG) https://orcid.org/0000-0002-5899-2052
• Raquel Cardoso de Melo-Minardi Universidade Federal de Minas Gerais (UFMG) https://orcid.org/0000-0001-5190-100X

DOI: https://doi.org/10.5753/bsb.2025.15148

Resumo

This work presents a machine learning approach for the virtual screening of histone deacetylase (HDAC) inhibitors, with a focus on the HDAC1 enzyme. Proposing a new structural signature based on the aromaticity of ligand atoms, the method models the protein-ligand binding region as a graph. Machine learning models were trained using this signature to distinguish ligands from decoys, achieving high accuracy. The methodology was applied to identify potential HDAC1 inhibitors from the T3DB database, resulting in the selection of compounds with binding potential. The results suggest that structural signatures are more efficient and computationally less expensive than molecular docking, thereby paving the way for the identification of HDAC inhibitors relevant to autism etiology studies.

Palavras-chave: Machine learning, Structural signatures, Histone deacetylase, Autism Spectrum Disorders

Referências

CIOLETTI, Alessandra et al. Using graph-based structural signatures and machine learning algorithms for molecular docking assessment of histone deacetylases and small ligands. In: BRAZILIAN SYMPOSIUM ON BIOINFORMATICS. Anais do XVII BSB. Vitória, Brazil: 2 dec. 2024.

LIM, Emilia et al. T3DB: a comprehensively annotated database of common toxins and their targets. Nucleic Acids Research, v. 38, n. Database issue, p. D781-786, jan. 2010.

MARIANO, Diego et al. A Computational Method to Propose Mutations in Enzymes Based on Structural Signature Variation (SSV). International Journal of Molecular Sciences, v. 20, n. 2, 15 jan. 2019.

MILLARD, Christopher J. et al. Class I HDACs share a common mechanism of regulation by inositol phosphates. Molecular Cell, v. 51, n. 1, p. 57–67, 11 jul. 2013.

PIRES, Douglas E. V. et al. aCSM: noise-free graph-based signatures to large-scale receptor-based ligand prediction. Bioinformatics (Oxford, England), v. 29, n. 7, p. 855–861, 1 apr. 2013.

SUN, Wenjie et al. Histone Acetylome-wide Association Study of Autism Spectrum Disorder. Cell, v. 167, n. 5, p. 1385-1397.e11, 17 nov. 2016.

WAYE, Mary M. Y.; CHENG, Ho Yu. Genetics and epigenetics of autism: A Review. Psychiatry and Clinical Neurosciences, v. 72, n. 4, p. 228–244, apr. 2018.

XIA, Jie et al. Comparative modeling and benchmarking data sets for human histone deacetylases and sirtuin families. Journal of Chemical Information and Modeling, v. 55, n. 2, p. 374–388, 23 feb. 2015.