Ritish Natesan
Research Mentor(s): Sriram Chandrasekaran
Mentor Department: Biomedical Engineering
Authors: Ritish Natesan, Margaret Reuter, Sriram Chandrasekaran
Session: Session 1 (9:00am – 9:50am)
Presentation Type: Oral
Abstract
The development of antibiotic resistance poses a growing threat to modern treatment of disease. To combat this, drug-drug combination therapies have been introduced which offer novel and effective therapeutic designs leveraging drug-drug synergies. Past computational and machine learning approaches have been introduced to predict synergistic combinations over a vast combinatorial space^, such as INDIGO*, which predicts synergies in E. Coli using chemogenomics data. Past literature* has suggested pharmacokinetic and pharmacodynamic interactions between classic antibiotics and traditional compounds found in common foods, plants, herbs, and flavorings, which create adverse and inhibitory effects. However, past literature* also suggests the potential for synergistic combinations between drugs and dietary compounds, along with the usage of computational and network models* to investigate them in precise manners. In discovering these combinations, new therapies can be designed which use fewer drugs and are more accessible for many communities. As such, our project seeks to elucidate potential synergies between antibiotics and traditional dietary compounds, such as those found in spices, herbs, and seeds. Our approach involves the use of the M2D2* machine learning pipeline, which will be used to predict compound-protein interactions between a set of antibiotics and a set of roughly 1000 dietary compounds and natural products, using SMILE structures to virtually represent them. This interaction data will be used by a second model to predict synergistic effects between the natural compounds and drugs. From there, we will perform a mechanistic analysis of metabolic pathways to discover mechanisms of action for potential synergies and antagonisms. With this computational approach, we hope to simulate thousands of combinations in an efficient manner, identifying potential combinations which could lead to novel therapy designs. Cantrell, J. , Chung, C. , Chandrasekaran, S. Machine learning to design antimicrobial combination therapies: promises and pitfalls. (2022) Chandrasekaran, S. et al. Chemogenomics and orthology-based design of antibiotic combination therapies. Molecular Systems Biology (2023) Chang, K., Choi, J. Food and Drug Interactions. J Lifestyle Med (2017) Huang, K. et al. DeepPurpose: a deep learning library for drug–target interaction prediction. Bioinformatics 1–6 (2020) Rigby, Use of Molecular Docking Simulations in Elucidating Synergistic, Additive, and/or Multi-Target (SAM) Effects of Herbal Medicines. 2024
