Examples of BERD-Supported Science
1. Statistical Methods for Comparative Effectiveness Research (CER) Studies: The BERD program director (Dr. Robert Lyles) conducts research to develop biostatistical methodology in the area of mismeasured data that is partially supported by a recent ARRA (RC4) grant administered by the National Institute of Nursing Research. This work develops methods for handling missing and misclassified data in CER studies and funds dissertation research for two Emory doctoral students (Ji Lin and Li Tang). The group has thus far published two manuscripts that cite the support of the ACTSI grant (Lyles and Lin, 2010; Lyles, Tang et al., in press). Importantly, the ARRA grant links Dr. Lyles and the doctoral students to Dr. Beau Bruce, a neuro-ophthalmologist and ACTSI investigator and former KL2 scholar. This research group is leveraging the methodology funding to carry out refinement and application of statistical approaches to the regression analysis of potentially misclassified binary ophthalmologic outcomes from Dr. Bruce’s NIH-funded K23 study of non-diluted ocular fundus photography in an emergency room setting. This is an example of a link initially forged under the ACTSI mechanism, in which other external funding enables the longer-term collaboration necessary to bring novel statistical methodology to bear on ACTSI-related research.
2. Novel Statistical Methods for High-Throughput Proteomics Experiments: A team of researchers led by Drs. John Hanfelt and Tianwei Yu (Emory BERD members) along with Dr. Junmin Peng (Emory Department of Human Genetics) developed the software toolkit SAPHIRRA (Statistical Analysis of Protein High-Throughput Robust Relative Abundances). This was accomplished with the assistance of Emory Biostatistics PhD graduate Dr. Sameera Wijayawardana, whose dissertation on methods for proteomics data analysis was completed in February 2011 under supervision of Drs. Hanfelt and Yu. SAPHIRRA provides a new set of statistical methods to analyze large proteomics data sets. Dr. Hanfelt’s work leveraged NIH support from the Emory Alzheimer’s Disease Research Center, a link facilitated by his BERD involvement. This demonstrates a significant methodological advance accelerated by CTSA funding.
3. Adaptation, Education and Motivation: Improving Evidence-based Medication Adherence: As an example of the leveraging of resources and integration across ACTSI institutions toward high-quality epidemiologic research, Dr. Robert Mayberry (MSM) serves as PI of this recently funded Agency for Healthcare Research and Quality (AHRQ). This project will adapt, customize, and deliver the content of comparative effectiveness research summary guides for oral and insulin medications aimed at enhancing prescribed medication adherence among a vulnerable population of low-income, urban African-American adults with type 2 diabetes. Dr. Lyles serves as co-investigator on the study and he and BERD provide senior-level biostatistical support.
4. Software for Liquid Chromatography-Mass Spectrometry (LC/MS) Data: LC/MS is the most widely used fingerprinting technique in metabolomics research. BERD member Dr. Tianwei Yu and colleagues (Yu et al., 2009) developed the software package apLCMS. This suite of methods allows reliable feature detection and recovery of low-intensity signals. Associated components facilitate examination of LC/MS spectra, which was previously done manually. The software package now serves as the standard LC/MS data processing pipeline in the clinical biomarkers lab at Emory headed by Dr. Dean Jones. Dr. Jones is involved in multiple ACTSI projects and Dr. Yu’s research in developing the software was funded under a joint ACTSI Pilot & Collaborative Translational & Clinical Studies (PiCoTraCS)/Emory University Research Committee (URC) pilot grant and under multiple program project grants awarded to the Jones lab. This demonstrates the leveraging of resources necessary to conceptualize and complete statistical developments with potential for significant impact on clinical and translational research.