Past Events

This seminar is a part 2 of 3 of the 2020 Biostatistics at the Frontier Seminar Series

Detecting biomarkers that are relevant to patients' survival outcome is crucial for precision medicine. Dimension reduction is key in the process. Although regularization methods have been used for dimension reduction, they cannot handle a large number of candidate biomarkers generated by modern bio-techniques. Variable screening, which has been widely used for handling exceedingly large numbers of variables, is however much underdeveloped for censored outcome data.

Data describing health outcomes of hidden populations and in low-resource areas are usually noisy and incomplete. In this talk, I will discuss two projects in such data-constrained settings.

Drs. Savage and McMaster will provide an update on the central NIH IRB, Non Human Subject Research (NHSR), and NCI Ethical Review Panel review processes for DCEG studies. All are highly encouraged to attend to receive clarifications on these processes.

It is increasingly common clinically for cancer specimens to be examined using techniques that identify somatic mutations. In principle these mutational profiles can be used to diagnose the tissue of origin, a critical task for the 3-5% of tumors that have an unknown primary site. Diagnosis of primary site is also critical for screening tests that employ circulating DNA.

This research lies in the development of statistical methods that can be used to inform clinical decision making. These methods aim to provide actionable, relevant information to patients and clinicians such as risk of a particular disease or response to a given treatment. In this seminar, Dr. Meisner will outline the broad themes of her research, discuss in detail two recent projects related to polygenic risk scores (PRS), and touch on possibilities for future work, including methods for PRS as treatment selection tools.

Seminar: Publicly available national survey data are useful for the evidence-based research to advance our understanding of important questions in the health and biomedical sciences. Dr. Yu discusses methods of appropriate variance estimation.

Genome-wide association studies (GWAS) have identified thousands of common genetic variants robustly associated with risk of cancer, yet the biological mechanisms underlying these associations remain largely unknown, in part because of confounding by local correlation among variants, and in part because most of these variants are located in non-coding regions with poorly-understood functionality. Methods that integrate GWAS results with auxiliary data can suggest biological hypotheses for further in silico, in vitro and in vivo studies.

In the early years of genome-wide association studies, data analysis primarily relied on fairly simplistic methods, such as running millions of univariate linear or logistic regressions, one for each genetic variant. Recently, however, as the sample sizes for some GWAS have become extremely large and various types of other genomic data have increasingly become available, analysis of such data has also become much more complex and statistically sophisticated.

Part 1 of 3 of the 2020 Biostatistics at the Frontier Seminar Series.