In-silico Model to Detect Antimicrobial Resistance in fungi known to cause infections in patients affected by SARC-CoV 19.
Project Duration: 10 months (2020-21)
Worked as: Project manager and Technical team lead
At: Pan Genomics
Funded by: Pan Genomics and Padmasiddha Healthcare
The primary goal of this project was to develop and implement a computational (in-silico) model to predict antimicrobial resistance (AMR) at the genomic level in fungal strains, specifically those affecting COVID-19 patients. Using Candida albicans as a model organism, the objective was to assess gene expression changes induced by a range of antifungal agents, identify resistant genetic markers, and create a predictive model for antimicrobial resistance mechanisms in clinical settings.
Approach and techniques used
We employed Candida albicans as a model organism to investigate its resistance to antifungal agents, including fluconazole, clotrimazole, econazole, terbinafine, and ketoconazole. The project combined experimental data on changes in lipid profiles, proteome characterization, and genome analysis to build an in-silico model for predicting resistance mechanisms. This model allowed us to assess resistance against nearly 10,000 molecules using bioinformatics algorithms.
In parallel, we explored antibiotic resistance in Salmonella typhi strains, testing the efficacy of antibiotics such as ampicillin, sulfonamides, ciprofloxacin, tetracycline, and penicillin. The approach included detailed gene expression analysis and computational simulations to determine resistance pathways in these bacterial strains.
Techniques Used:
-
Genomic sequencing and annotation
-
Proteome analysis using mass spectrometry and bioinformatics tools
-
In-silico simulations for AMR prediction in a software we developed
-
Gene expression profiling
-
Antibiotic susceptibility testing
-
Molecular docking and modeling
-
Software development
Output and Impact
The computational model successfully predicted resistance mechanisms for Candida albicans, offering a significant reduction in the time required for AMR assessment. The model was extended to other pathogenic fungi, leading to substantial efficiency improvements in clinical settings by minimizing direct pathogen interaction and resource expenditure. However, the model did not consistently produce accurate predictions for Salmonella typhi across all sub-strains, highlighting the need for further refinement in bacterial resistance analysis.
SIGNIFICANCE & IMPACT ON SOCIETY
This project introduced a pioneering computational approach to AMR detection, providing a framework that could be widely adopted in clinical and research settings. By reducing the time and effort required for traditional resistance testing, the model enhanced the speed and accuracy of diagnosis and treatment plans, particularly during the COVID-19 pandemic when timely decisions were critical. The methodology also provided a safer alternative by reducing the need for direct handling of pathogenic strains in the lab, making it a valuable tool in future public health crises.
