- Data Scientist at Facebook
Researchers in the social and computer sciences have been studying how information diffuses through social networks, and this research has meaningful implications for many areas of public health. In this talk I will present research on the factors and consequences of information sharing in social networks, including the structure of the networks, the content of the message, and the interactions between the two. Additionally, I will talk about how health researchers can leverage social media and the internet more broadly to create effective interventions.
Winter Mason is a Data Scientist at Facebook, whose research bridges psychology and computer science, focusing on social networks, social media, and crowdsourcing. Prior to working at Facebook, he received his Ph.D. in social psychology and cognitive science in 2007 from Indiana University, went on to work as a visiting scientist with the Human Social Dynamics group at Yahoo! Research and then as an assistant professor in the Howe School of Technology Management at Stevens Institute of Technology.
- Director, Vanderbilt Genetics Institute
Director, Division of Genetic Medicine
Vanderbilt Brain Institute
- Vanderbilt University
The phenome, obtained via large-scale electronic health records (HER), has been the subject of in-depth investigations in a number of contexts but has not yet been broadly integrated with -omics data. We have developed an approach for integrating transcriptome and genome variation data in a reference sample such as the data from the Genotype Tissue Expression (GTEx) project in which we create SNP-based prediction equations for transcript levels for each gene in each tissue (Gamazon et al, Nat Genet 2015). With this Predicted Expression Scanning (PrediXcan) approach, we can then calculate for all individuals in a biobank with genome-wide data (GWAS or whole genome sequencing) the genetically regulated expression (GReX) of each gene for each tissue , and conduct association studies between GReX for each gene and PheWAS codes, a mapping of electronic health records billing codes to related subgroups (insuring sufficient numbers of individuals in each subgrouping). The result is a catalog characterizing the human medical phenotypes associated with alterations in the expression of each gene. We have conducted such analyses in BioVU, the biobank at Vanderbilt University, having DNA on more than 210,000 individuals linked to de-identified and continuously updated EHR going back on average 10-15 years, and for a subset of individuals up to 30 years. Because our signals come at the level of the gene and have an easy-to-interpret direction of effect, it is more straightforward to use results of these studies to characterize gene-phenome relationships. I will characterize ³big picture² observations from the catalog to date, as well as providing examples of novel discoveries that have been validated in model systems studies. While we are still early in the development of this catalog, we will have more than 100,000 subjects completed by the first quarter of 2017, and ultimately expect to develop a comprehensive catalog of gene-phenome relationships.
Nancy J. Cox is a quantitative human geneticist who has worked for more than 30 years to identify and characterize the genetic component to common diseases and complex traits including pharmacogenomics phenotypes. Current research is focused on data integration approaches, including genomics, transcriptomics, and large-scale mining of EMR data. Dr. Cox earned a BS in biology from the University of Notre Dame in 1978, a PhD in human genetics from Yale University in 1982, and did post-doctoral research at Washington University and the University of Pennsylvania before joining the University of Chicago in 1987, where she spent 28 years as a faculty member in the Depts of Medicine and Human Genetics. In 2015, she joined Vanderbilt University as the first of the Vanderbilt Genetics Institute.