DATA ANALYSIS AND STATISTICAL PROGRAMING SUPPORT (BASE CONTRACT) COMPANY: PROSPECTIVE GROUP

点击翻译按钮获取中文摘要

The Division’s research often involves targeted population subgroups including couples of reproductive age of which some may be planning pregnancies, women with low and high-risk pregnancies, or children and adolescents with risky behaviors along. Sampling frameworks utilized by the Division typically include population-based strategies (e.g., registries, marketing databases) and clinically based sampling (e.g., billing and clinic records, surgical schedules). To the extent feasible, the referent population is delineated for all sampling frameworks including those implemented within clinical facilities. The Division’s research is often conducted with nongovernmental investigators from either schools of medicine and biomedical sciences or public health via research and development contracts. All Division research is highly collaborative and trans-disciplinary reflecting the complexity of our research questions, novel study protocols and our passion for answering critical data gaps with the ultimate goal of promoting the health and well-being of populations. Of note is the dual publication track record for much of the Division’s research. For example, biobehavioral and epidemiological investigators will publish the results in their respective journals, while the biostatistical investigators will use this work to motivate original methods research and to publish methods papers in statistical journals. In addition, Division investigators publish their research in subject matter specialty and population health journals. The Division’s research includes both observational and experimental study designs, with most research being prospective in nature and with longitudinal data capture including the collection of biospecimens and in some studies, imaging data (e.g., digital video recording, pregnancy ultrasounds). Of note is the hierarchical data structure underlying much of our work either from the use of triads/diads or genome wide analytic studies (GWAS) or multiscale data from study participants (e.g., day, cycle, woman and couple level data collection for fecundity and fertility research). The highly timed and conditional nature of human reproduction and development is well suited for statistical methods such as joint modeling. Examples of prospective cohort studies with longitudinal measurement and biospecimens and a hierarchical data structure include: the NICHD Fetal Growth Studies (FGS), Longitudinal Investigation of Fertility & Environment (LIFE Study), Next Generation Health Study (NEXT), and the Diabetes and Women’s Health Study. Examples of studies with high dimensional data requiring GWAS and EWAS techniques include: Endometriosis: Natural History, Diagnosis, and Outcomes (ENDO) and genetic determinants of birth defects. Examples of our randomized trials include: Cultivating Healthy Eating in Families of Youth with Type 1 Diabetes (CHEF Trial), Family Management of Diabetes (FMOD); The Teen Passenger Simulation Study; Effects of Aspirin on Gestation and Reproduction Trial (EAGeR), and the recently completed Folic Acid and Zinc Supplementation Trial (FAZST). A description of all studies with publications to date can be found at the Division’s website https://www.nichd.nih.gov/about/org/diphr/Pages/default. Our current research portfolio has approximately 20 large population-based studies ranging from randomized intervention trials (n=260) to large cohort studies (n=3600). Investigators in the Biostatistics and Bioinformatics Branch lead the development of the analytic plans for Division research in collaboration with investigators from other branches. With guidance from Division Investigators, the Contract staff will implement the analytic plans. Objective/Specific Aims The purpose of this contract is to provide program support for Epidemiologists, Biostatisticians and Behavioral Scientists in the Division to enable their effectiveness. The proposed work includes program support for statistical programming, data collection, management, analysis and reporting. • Perform programming with high level knowledge of SAS and R; • Use, when available, public free-ware for data management and analysis; • Condense, merge and reformat data into files that are appropriate for data analysis (e.g., R,SAS, Stata, SPSS, MPlus, PLINK, Bioconductor); • Condense, merge and reformat data into files that are appropriate for data sharing; • Combine original data that have formats such as ASCII, Excel, data from labs, including genomics and proteomics testing, image data, as well as other formats; • Use a facility with cloud computing and data storage; • Create complex variables using longitudinal data; • Handle genomics data such as metabolomics and proteomics, in situations where the data is measured longitudinally; • Store and retrieve “omics” data studies with a very large number of participants and store longitudinal metabolomic profiles; • Store and retrieve sonographic images from complex multicenter longitudinal studies; • Store and retrieve longitudinal ultrasound images in studies with a very large number of participants; • Condense, merge and format data collected from genetic studies for analytic programs such as R, SAS, Stata, Bioconductor, and Plink. Original data have formats such as ASCII, Excel, data from labs, including genomics and proteomics testing, image data, and other formats. Emphasis shall be placed on using publicly available software for data storage; • Create complex variables using longitudinal data; • Create unique data bases and perform error checking, cleaning and running summaries. Examples include quality control for large longitudinal cohort studies to smaller intervention studies. All studies have complex data structures including, but not limited to, imaging, high-dimensional biomarkers, accelerometer and sleep measurements, driving performance as measured from simulations and naturalistic driving studies; • Prepare de-identified datasets and documentation specifically for datasharing; • Interact with web sites used for sharing data and information. Types of web sites include: Web sites for disseminating information to the public as in our CheckPoints site, https://checkpoints.nichd.nih.gov; Web sites for collaborative work; Web sites for sharing data with the research public; example https://brads.nichd.nih.gov; • Create or update websites for studies or datasharing with password-protected areas to facilitate secure sharing of manuscripts, access to data, and sharing of findings, as needed; • Create phone applications for research studies, as needed.