PseudoResistance DB: A new Database of antibiotics related to Pseudomonas aeruginosa antibiotic resistance
Description
This research addresses the pressing issue of antibiotic resistance, a global health challenge that undermines the efficacy of treatments against infectious diseases. Focusing on Pseudomonas aeruginosa—a Gram-negative bacterium known for causing opportunistic infections—this study emphasizes its prioritization by the World Health Organization (WHO) as a critical-level pathogen requiring new therapeutic approaches. To identify antibiotics associated with P. aeruginosa, the study employed text mining techniques on the Scielo database. The resulting dataset comprises 98 antibiotics, each documented with detailed textual information and referencing data. Additionally, the dataset includes structural files of the antibiotics in several formats suitable for computational modeling and simulations. These formats encompass Protein Data Bank, Partial Charge & Atom Type (PDBQT), Simplified Molecular Input Line Entry System (SMI), IUPAC International Chemical Identifier (INCHI), Molecular Design Limited Molfile (MOL2), Structure-Data File (SDF), Chemical Markup Language (CML), Cartesian Coordinates File (XYZ), Scalable Vector Graphics (SVG), Molecular File (MOL) and Protein Data Bank (PDB) files, with molecular models generated via OpenBabel to facilitate advanced studies in drug development and resistance mechanisms.
Files
Steps to reproduce
This study developed a comprehensive database of antibiotics interacting with RND resistance systems in *Pseudomonas aeruginosa* (PA), using a systematic, reproducible approach. First, keywords relevant to antibiotic resistance in PA were identified from the literature. These keywords guided searches in the SciELO database, yielding references in BibTeX format. A Python script merged these files, removed duplicates, and extracted article links into a CSV. Using the requests and BeautifulSoup libraries, automated searches retrieved article content, which was scanned for antibiotic mentions. Terms meeting specific criteria were retained, while others—such as non-molecular classes and complex peptides—were excluded. This filtering narrowed the database to 98 antibiotics across 26 classes. Finally, using Open Babel, molecular data was converted into multiple formats, ensuring compatibility for computational analysis and future research on PA antibiotic resistance.