MD Jobayer Hossain, PhD

Biography

I am a Senior Research Scientist in the Department of Biomedical Research and the Director of the Biostatistics Program at Nemours. Additionally, I serve as a Research Professor at the Sidney Kimmel Medical College of Thomas Jefferson University and as adjunct faculty at the Center for Bioinformatics & Computational Biology, as well as the Department of Applied Economics and Statistics at the University of Delaware. My research interests focus on the innovative application of statistical, data science, and machine learning methods to ensure the effectiveness of clinical, epidemiological, and public health research studies. As the director of the Biostatistics Program, I lead core activities of the theoretical development and innovative application of statistical methods and provide statistical direction and technical oversight for projects of interdisciplinary collaborations that require statistical input. Specifically, I provide expert-level inputs to research projects of various sizes, scopes, phases, and levels of funding, and provide statistical support to clinical researchers in designing clinical trials, pre-clinical, epidemiological, and survey studies; power and sample size calculation; randomization; data management; and data analysis; and in authoring the statistical section in research protocols, manuscripts, abstracts, and presentations. I have broad research collaborations with clinicians of many different disciplines, and I have led several studies as principal investigator (PI) including a COBRE pilot grant to investigate early childhood growth patterns using Nemours electronic health records. I also provide training and education on statistics and analytical software skills at Nemours and affiliated institutions. In my graduate course, Electronic Health Records (EHRs) Data Science, students in data science, bioinformatics, and computational biology master healthcare databases and data sources. The course covers biomedical data science, EHR structure and management, and big data analytics. Students learn database systems, models, and data structures, importing data from spreadsheets, Google Sheets, images, and webpages. The course emphasizes data analysis through statistical methods and machine learning, along with data visualization. Additionally, we cover text mining and natural language processing (NLP) for text data analysis. My own and collaborative work have produced illustrative methodological innovation and explored etiologic pathways and health trends in diverse pediatric populations to improve the health of children affected by a broad range of diseases, disorders, and health conditions, including pediatric obesity, asthma, allergy, diabetes, and cancer. In addition to routine statistical modeling of data, my current research interests include tracking changes in health trajectories over time (e.g., childhood growth trajectories); identifying refraction of health trajectories around critical events (e.g., medical diagnosis); recognizing hidden patterns in longitudinal changes (e.g., growth patterns) or in correlated variables of overlapping as well as contrasting characteristics in cross-sectional data; forming and predicting classes of distinct patterns; and identifying individual- as well as community-level demographic, clinical and other features responsible for diverse trajectories. In my career as a statistician, I have extensive experience applying statistical methodology in clinical, epidemiological, pharmaceutical, and financial settings. My work includes cross-sectional, longitudinal, case-control, and survey data analyses; linear and nonlinear parametric and non-parametric regressions; mixed and random effects, survival analysis, time series analysis; machine learning and data mining methods; regularization and selection techniques; and cross-validation. My clinical trial experience (phases I-IV) in various therapeutic areas aids in interacting with clinicians. I guide junior statisticians, teach researchers, medical fellows, and research associates at Nemours, and occasionally teach statistics courses at the University of Delaware. My graduate course work covered a wide range of areas in statistics (classical and Bayesian), biostatistics, population science (demography), data science and machine learning, mathematics (numerical mathematics and optimization theory), probability (measure theoric, stochastic, and inferential statistics), computer sciences, and economics. I am a lead or co-author for 150 published articles and more than 260 published abstracts and presentations. I am actively involved in scientific, educational, and research activities. My academic training and extensive work experience in statistics, epidemiology, and data science have prepared me to be an effective research and teaching faculty in relevant fields.