# Toxicity assessment of doxycycline-aided artificial intelligence-assisted drug design targeting candidate 16S rRNA methyltransferase gene

## TL;DR

This paper explores using AI to design drugs targeting a gene linked to antibiotic resistance, with doxycycline showing promise in early tests.

## Contribution

A novel AI-driven approach for drug design targeting the 16S rRNA methyltransferase gene, combined with toxicity assessment of doxycycline.

## Key findings

- AI-designed doxycycline shows non-toxic properties and high absorption through the blood-brain barrier.
- The AI-designed drug exhibits strong docking affinity with the 16S rRNA methyltransferase protein (-7.6 kcal/mol binding energy).
- In silico studies suggest therapeutic potential but emphasize the need for in vivo trials to confirm results.

## Abstract

The misfunction of the protein 16SrRNA methyltransferase can result in Urinary tract infections (UTI), Gastrointestinal (GI) infections, sepsis, pneumonia, and wound infections; various tactics are used to lessen the fatal consequences. It confers resistance to aminoglycoside medications, which complicates the treatment of infections caused by these bacteria. Innovative methods are desperately needed to stop these diseases from spreading because there are no reliable medical therapies available.

Herein, we aim to evaluate Doxycycline’s Role in AI-Driven Drug Design and identification of effective inhibitors targeting the 16S rRNA methyltransferase gene. Additionally, to investigate the toxicological profiles of designed drug through AI approach for advancement in medical sciences.

Methodology involves, selection of three effective de novo medicinal compounds that target the 16SrRNA methyltransferase protein for designing an AI driven drug. Multiple in silico tools were used for designing AI based drug includes: Expasy for protein annotation, ProtParam to calculate physiochemical parameters, SWISS-MODEL to estimate the 3D structure, and UniProt to generate the 16SrRNA methyltransferase protein sequence. An adequate foundation for the development and validation of AI-designed phytochemical medicines for infections is provided by quality assessment, binding site prediction, drug design with WADDAICA, toxicity screening, ADMET evaluation, and docking analysis with CB-dock.

Comprehensive pharmacokinetic and toxicology analyses confirm that the AI-designed doxycycline exhibits a non-toxic character, with particularly high absorption through the blood-brain barrier. Furthermore, the AI-designed doxycycline docked complex demonstrates a strong docking affinity with the 16S rRNA methyltransferase protein, showing a binding energy of approximately − 7.6 kcal/mol, suggesting significant therapeutic potential.

Even though the in silico studies show efficacy and safety, still there is need of in vivo trials to investigate the hidden medical aspects. By addressing existing constraints, presenting a non-invasive approach to infections, and providing viable substitutes for traditional surgical procedures, this work considerably expands the knowledge about newer methods and also helps to understand deep insights of dug design mechanism for treatment.