Patient digital twin for peptide muscle relaxant ativity control

Physiologically based pharmacokinetic (PBPK) modeling serves as an important tool in drug development and personalized therapy. It enables the integration of individual patient physiological characteristics to predict drug distribution across tissues, which is particularly relevant for chronic diseases, impaired liver and kidney function, as well as for off-label application scenarios or evaluation of new compounds.

This work presents an interactive PBPK model that describes drug distribution across 13 major body compartments: blood, liver, kidneys, lungs, heart, muscle and adipose tissue, brain, bones, gut, spleen, skin, and other tissues. The model is based on systems of ordinary differential equations (ODEs). Key pharmacokinetic parameters are incorporated in the calculations: some are determined experimentally (e.g., logP, unbound drug fraction, hepatic clearance), while others are obtained from literature sources (e.g., pKa).

A distinctive feature of this development is its complete implementation within the open-source R environment using the deSolve package for numerical integration of equations. This approach ensures full transparency, reproducibility of results, and easy integration with other statistical and bioinformatic tools. The model is integrated with an intuitive Shiny/plotly web interface, allowing flexible adjustment of administration route, organ clearance, dosing regimen, and frequency.

As a case study, simulations were performed for azemiopsin - a peptide neurotoxin with muscle relaxant properties. The model demonstrated pronounced drug accumulation in muscle tissue with rapid clearance from systemic circulation. An additional module simulated antibody-antigen interaction, which is considered and modeled within the framework of the drug-drug interaction (DDI) concept. This demonstrates the platform's capability to analyze complex pharmacokinetic scenarios and various dosing regimens.

The PBPK model shows high flexibility and practical value. It can be used for planning preclinical studies, optimizing experimental protocols, refining ADME profiles, and predicting individual body responses to drugs, thereby creating a foundation for personalized pharmacotherapy.