Presented By: Biomedical Engineering
BME Master's Thesis Defense - Suraj Jaipalli
Discovering Synergistic and Antagonistic Drug Combinations for Mycobacterium Tuberculosis
With the rise of antibiotic resistance, treatments involving common frontline TB drugs have been rendered insufficient, especially in treating multidrug resistant (MDR) and extremely drug resistant (XDR) strains of TB. As new anti-TB therapies are beginning to emerge, optimizing specific combinations of individual agents into effective and safe regimens remains a significant challenge. To address this problem, we have developed a in-silico computational tool—Inferring Drug Interactions using chemo-Genomics and Orthology for MTB (INDIGO-MTB). INDIGO-MTB leverages high resolution MTB transcriptomic profiles and TB drug interaction data to predict synergy/antagonism of TB drug regimens with high accuracy. INDIGO-MTB uses the random forests machine learning algorithm to train its predictive model, which can be used to make predictions for synergy of novel TB drug regimens. INDIGO-MTB predictions correlated well with corresponding in-vitro drug interaction validation scores. The in-vitro INDIGO-MTB interaction scores were also predictive of the percentage of patients with negative sputum cultures after 8 weeks in clinical trials for 58 TB drug regimens. We hope INDIGO-MTB can be used by clinicians and researchers to quickly assess the likelihood of success of new TB drug combinations using publicly available data.
Chair: Sriram Chandrasekaran, Ph.D.
With the rise of antibiotic resistance, treatments involving common frontline TB drugs have been rendered insufficient, especially in treating multidrug resistant (MDR) and extremely drug resistant (XDR) strains of TB. As new anti-TB therapies are beginning to emerge, optimizing specific combinations of individual agents into effective and safe regimens remains a significant challenge. To address this problem, we have developed a in-silico computational tool—Inferring Drug Interactions using chemo-Genomics and Orthology for MTB (INDIGO-MTB). INDIGO-MTB leverages high resolution MTB transcriptomic profiles and TB drug interaction data to predict synergy/antagonism of TB drug regimens with high accuracy. INDIGO-MTB uses the random forests machine learning algorithm to train its predictive model, which can be used to make predictions for synergy of novel TB drug regimens. INDIGO-MTB predictions correlated well with corresponding in-vitro drug interaction validation scores. The in-vitro INDIGO-MTB interaction scores were also predictive of the percentage of patients with negative sputum cultures after 8 weeks in clinical trials for 58 TB drug regimens. We hope INDIGO-MTB can be used by clinicians and researchers to quickly assess the likelihood of success of new TB drug combinations using publicly available data.
Chair: Sriram Chandrasekaran, Ph.D.
Explore Similar Events
-
Loading Similar Events...