Danping Liu, Ph.D.
Biostatistics & Bioinformatics Branch
Statistical Methods in Longitudinal Biomarker Research with Missing Data
Individualized risk prediction is particularly attractive to patients, health care providers and policy makers. At the population level, accurate prediction of disease helps target interventions and treatments to the most vulnerable groups. Biomarker research has attracted much attention in the diagnosis or prediction of disease outcomes, especially in studies of pregnancy and child development outcomes. My research interest lies in developing new statistical methods to evaluate and to combine biomarkers. Particularly, I am interested in the following topics:
Developing methods to combine longitudinal biomarkers.
A common practice in biomarker research is to take multiple measurements over time. The prediction accuracy could be substantially improved by accounting for the longitudinal trajectories of the biomarkers and their correlation structures. In addition, individual risk calculators could be developed based on the biomarker combination rule. I have been working on methods for effectively combining longitudinal biomarkers in predicting pregnancy complications.
Developing model selection techniques for combining high-dimensional longitudinal biomarkers.
Including more biomarkers into the combination does not always improve the prediction accuracy. Selection of markers is a crucial step, especially when the number of candidate markers is large. Marker selection involves several aspects: selecting markers with high classification power, selecting the important time points for making observations, and selecting subgroups of patients for enhanced prediction. I am interested in developing the model selection techniques in the longitudinal biomarker combination framework.
Missing data problems in biomarker evaluation and combination.
Incomplete observations are common in most longitudinal studies, but this problem has not been studied in the context of biomarker combination. Both the biomarker and the outcome could be missing due to loss of follow-up. The missingness might be informative if the dropout process depends on a subject’s unobserved characteristics. For example, when predicting preterm birth, the subjects with early preterm tend to have a shorter sequence of biomarker evaluations. My interest lies in developing efficient and robust techniques to correct for the bias resulted from ignoring the missing data.
Danping Liu, Ph.D., joined the Biostatistics and Bioinformatics Branch as a principle investigator in 2012. He received his Ph.D. in biostatistics from the University of Washington in 2010. His research interests include biomarkers and medical diagnosis, analysis of longitudinal data, missing data methodologies, and semiparametric inference.
Fulton KA, Liu D, Haynie DL, Albert PS. MIXED MODEL AND ESTIMATING EQUATION APPROACHES FOR ZERO INFLATION IN CLUSTERED BINARY RESPONSE DATA WITH APPLICATION TO A DATING VIOLENCE STUDY. Ann Appl Stat. 2015;9(1):275-299.
Foster JC, Liu D, Albert PS, Liu A. Identifying subgroups of enhanced predictive accuracy from longitudinal biomarker data using tree-based approaches: applications to fetal growth. J R Stat Soc Ser A Stat Soc. 2017;180(1):247-261.
Liu D, Zhou XH. Covariate adjustment in estimating the area under ROC curve with partially missing gold standard. Biometrics. 2013;69(1):91-100.
Liu D, Albert PS. Combination of longitudinal biomarkers in predicting binary events. Biostatistics. 2014;15(4):706-18.
Liu D, Yeung EH, McLain AC, Xie Y, Buck Louis GM, Sundaram R. A Two-Step Approach for Analysis of Nonignorable Missing Outcomes in Longitudinal Regression: an Application to Upstate KIDS Study. Paediatr Perinat Epidemiol. 2017.
Related Scientific Focus Areas
Social and Behavioral Sciences
This page was last updated on February 23rd, 2018