Computational Analysis of Host, Pathogen and Drug Interactions

Technology Overview

Dr. Shuyi Ma

An estimated 1.8 billion people worldwide are infected with the bacterium that causes tuberculosis (TB). Global control of TB is hampered by widespread resistance of Mycobacterium tuberculosis to frontline antibiotic treatments. Combination therapy with multiple antibiotics can be an effective strategy against drug-resistant bacteria because a bacterial mutation that confers resistance to one antibiotic might not protect against a drug with a different mechanism of action.

An in-depth understanding of interactions between drugs and their target pathogens, along with knowledge about drug-drug synergies or interferences are valuable resources for designing effective combination therapies. Adding information about host-drug interactions can help prevent adverse reactions and individualize medication regimens for maximum effectiveness. Comprehensive data about the interactions among drugs, pathogens and hosts can guide the selection of effective combination therapies and support the discovery of new therapeutic targets and the identification of additional indications for already-approved drugs.

Models and workflows for safe, effective new drugs and drug combinations

Systems biologist Shuyi Ma, PhD, integrates computational and experimental workflows for high-throughput analysis of large datasets on gene regulation, gene expression, metabolic activity and phenotypic fitness. Dr. Ma and her team create machine learning models and structure the resulting output information into testable hypotheses to drive drug discovery.

Dr. Ma’s team applies their expertise to determining and quantifying the networks that describe interactions between pathogens and host cells at multiple stages of infection. This research has identified the conditions, along with the host genes and pathogen genes, that act at tipping points that shift an infection either toward more serious illness or toward pathogen clearance and host recovery. Computational models and workflow pipelines developed by the Ma Lab have organized data from prokaryotic and eukaryotic RNA-seq into gene regulation and expression networks that reveal drug-pathogen-host relationships. These results from the research team facilitate the development of both pathogen-targeting and host-directed new therapies.

For example, Dr. Ma’s Lab created and validated an in silico-guided platform that used transcriptome data from M. tuberculosis treated with a single antibiotic to predict drugs that, if used in combination, would be an effective multidrug therapy against TB. In vitro analyses verified many of the predicted complementary and synergistic drug relationships. These potential combination therapies are ready for further testing and validation.

The Ma Lab can design systems to analyze a variety of data types. Using microscopy images from an animal model, the research team developed a platform for reporting the cell-level physiological effects of medications on hair cells inside the ear that are crucial for hearing. This research predicted drug interactions that either caused ototoxicity or protected hair cells from this common adverse medication effect that can result in hearing problems, ringing in the ears and disrupted balance.

Dr. Ma is interested in industry partnerships that use her team’s skills and experience in applying analytic methods including artificial intelligence (AI) modeling to large datasets. The Ma Lab aims to apply their expertise in network and systems biology and computational methods including AI to develop research tools and clinical products that improve treatments for infectious diseases, particularly TB.

Stage of Development

• Preclinical in silico
• Preclinical in vitro
• Preclinical in vivo

Partnering Opportunities

• Collaborative research and development
• Sponsored research agreement
• Licensing agreement

Learn More

• Shuyi Ma, PhD
• Ma Lab

Publications

1. Ivie JJ, Stull S, Bustad E … Ma S. Modulation of ferroptosis during early Mycobacterium tuberculosis infection contributes to Beijing Lineage Strain SA161 virulence. bioRxiv [Preprint]. 2026.05.29.728786.
2. Griebel BT, Ma S. MetworkPy: a Python package for graph- and information-theoretic investigation of metabolic networks. bioRxiv [Preprint]. 2026.05.26.727944.
3. Bustad E, Petry E, Gu O … Ma S. Predicting fitness in Mycobacterium tuberculosis with transcriptional regulatory network-informed interpretable machine learning. Front Tuberc. 2025;3:1500899.
4. Bustad E, Mudrock E, Nilles EM, Ma S. In vivo screening for toxicity-modulating drug interactions identifies antagonism that protects against ototoxicity in zebrafish. Front Pharmacol. 2024;15:1363545.
5. Ofori-Anyinam B, Hamblin M, Coldren ML … Ma S, et al. Catalase activity deficiency sensitizes multidrug-resistant Mycobacterium tuberculosis to the ATP synthase inhibitor bedaquiline. Nat Commun. 2024;15(1):9792.
6. Ma S, Morrison R, Hobbs SJ, et al. Transcriptional regulator-induced phenotype screen reveals drug potentiators in Mycobacterium tuberculosis. Nat Microbiol. 2021;6(1):44-50.

Last updated August 2026