A Structural Bioinformatics Framework for Prioritizing Ph-Sensitive Proteins From 3D Structural Features Publisher Pubmed



Arsanjani A A ; Malekshahi Z V ; Mosayyebi B ; Negahdari B ; Amirlou M ; Khavari F ; Faradonbeh D R
Source: BMC Bioinformatics Published:2026

Abstract

Background: Acidic extracellular pH is a defining feature of many solid tumors and can influence the structural stability and activity of proteins involved in cancer progression. However, no unified computational framework exists for systematically assessing the pH-sensitivity of proteins using structural information. Methods: We developed a structural bioinformatics framework that quantifies pH-sensitivity by integrating five protonation-relevant descriptors derived from protein sequences and 3D structures: total histidine count, histidine proportion, ionizable residues with predicted pKa values in acidic ranges, solvent-exposed histidines, and helix-associated histidines. pKa values were obtained using PROPKA 3.0, and all feature extraction steps were implemented through custom Python scripts. The framework was applied to cancer-associated secretory and membrane proteins to prioritize those whose structures may be particularly susceptible to acidic tumor conditions. Results: The framework effectively ranked proteins according to their structural susceptibility to acidic pH and revealed distinct distributions of pH-sensitive features across secretory and membrane classes. Several high-scoring proteins, including ITGB8, LGR5, FZD7, HLA-E, ANGPTL2, and BST2 emerged as candidates potentially affected by protonation-linked perturbations in acidic tumor microenvironments. Enrichment analyses of ranked subsets demonstrated how score-derived protein groups can support downstream biological interpretation. Conclusion: This study presents a reproducible and extensible structural scoring framework and applies it to the systematic prioritization of cancer-associated proteins potentially vulnerable to acidic tumor pH. By combining 3D structural descriptors with protonation predictions, the approach offers a computational tool for investigating pH-dependent protein behavior in cancer biology and other contexts where local acidity shapes molecular function. © The Author(s) 2026.