Journal of Clinical & Medical Genomics

ISSN: 2472-128X

Open Access

Paramita Saha Chaudhuri

Paramita Saha Chaudhuri

Paramita Saha Chaudhuri
Assistant Professor, Department of Biostatistics and Bioinformatics
Duke University School of Medicine, USA


I received my PhD in Biostatistics from Department of Biostatistics, University of Washington, Seattle in 2009. I am an Assistant Professor in the Department of Biostatistics and Bioinformatics at Duke University. I was a Research Fellow at the Biostatistics Branch of The National Institute of Environmental Health Sciences (NIH) from June, 2009 - Sep, 2011.

Research Interest

Keywords: Time-dependent ROC, ROC for Cox model, AUC, C-index, competing risks, classification, prediction, accuracy, diagnostic, survival, longitudinal; group testing, specimen pooling, epidemiology, methods for observational studies, efficient study design for expensive biomarker, efficient use of archived specimen, DNA pooling; statistical software for time-dependent ROC.

Time-dependent Predictive Accuracy: I am interested in developing new methodology to assess predictive accuracy of a marker or a risk-score or a set of covariates to predict a time-to-event or survival outcome. Standard diagnostic accuracy summaries like True and False Positive Proportions and ROC curves were extended to accommodate censored survival data and I hope to further extend this new area of research.

Pooled Exposure Analysis for Effect Assessment: I am also interested in efficient (cost-effective) study designs and how they can be employed in risk assessment. Currently I am researching pooled specimen analysis as a cost-effective alternative to standard analysis for estimating risk parameters.

Gene by Environment Interaction in Longitudinal setting: I recently got interested in estimation of GxE interaction based on longitudinal data. I am learning about powerful tests of interaction based on reduced models, such as Tukey's 1-df model.

Other: Statistical methods for observational and epidemiological studies and in particular, methods for survival analysis, classification and prediction, longitudinal data are also of interest.

arrow_upward arrow_upward