Embedded Files
This work reflects a collaborative effort while at Auburn University. I work closely with various graduate researchers and faculty in chemistry, pharmacy, and veterinary sciences through joint projects applying advanced statistical and machine learning methods to real scientific and biomedical data
Chemistry
Work with the chemistry department to develop a machine-learning pipeline for analyzing LC-IM-MS data, enabling robust feature extraction, visualization, and classification across bacterial strains. The framework clusters peaks globally across replicates using mass, retention time, and drift time tolerances, applies voting to retain discriminative features, and integrates PCA, hierarchical clustering, and Random Forest models to identify informative molecular signatures and improve strain-level discrimination
In progress: Kimberly Y. Kartowikromo, Michael Zirpoli, Jingyi Zheng, and Ahmed M. Hamid, “Multidimensional Ion Mobility Mass Spectrometry for Structural Characterization of Alzheimer’s Disease Biomarkers”
Orobola E. Olajide, Michael Zirpoli, Kimberly Y. Kartowikromo, Jingyi Zheng, and Ahmed M. Hamid, “Discrimination of Common E. coli Strains in Urine by Liquid Chromatography–Ion Mobility–Tandem Mass Spectrometry and Machine Learning”, Journal of the American Society for Mass Spectrometry, 2024
Pharmacy
Developed a biochemometric analysis pipeline using multivariate statistics and machine learning, including PCA, hierarchical clustering, DBSCAN outlier detection, and Random Forest classification, to characterize chemical and bioactivity patterns across extract fractions, enabling the identification of compounds associated with biological activity
Submitted: Zarna Raichura, Michael Zirpoli, Rebecca Paul-Hamby, Chris Harrilal, Brooke Dugan, Kate Givens, Jingyi Zheng, Angela I Calderon, “Integrated Assessment of Cytochrome P450 Inhibition by Acai and Mass Spectrometry-Based Identification of CYP Inhibitory Constituents Using Biochemometric Approaches”, Drug Metabolism and Disposition, 2026
Veterinary
Developed data-driven prognostic models for veterinary clinical datasets using univariate and multivariate logistic regression, including Firth’s penalized methods for small-sample and separation settings, and Random Forest–based imputation to identify predictors of survival, discharge, and readmission
Submitted: Hannah Maxwell, Michael Zirpoli, Dane W. Schwartz, Jessica Rush, Thomas Passler, Manuel F. Chamorro, “Prognostic Indicators for Survival in Goats with Toxic Mastitis: A Retrospective Study”, The Canadian Veterinary Journal, 2025
Submitted: Grace Butler, Michael Zirpoli, Dane W. Schwartz, Thomas Passler, Miguel Saucedo, Jenna E. Bayne, Jenna Stockler, Jessica Rush, Chance Armstrong, Katelyn Waters, Ricardo Stockler, Manuel F. Chamorro, “Prognostic Indicators for Survival and Readmission in Cattle Undergoing Umbilical Surgery”, The Canadian Veterinary Journal, 2025
Page updated
Google Sites
Report abuse