Quantitative Methods for HIV/AIDS Research Spurred by recent technological advances, HIV/AIDS research is generating increasingly large and complex data sets. For example, the data collected in a single vaccine trial might include viral sequence, antibody binding and functional responses to multiple epitopes, and several flow or mass cytometry panels characterizing cellular phenotype and antigen-specific binding, activation and function. New genomic technologies such as Fluidigm or Droplet based single-cell RNA sequencing represent orders of magnitude increases in data complexity, and will push HIV/AIDS research even further into the realm of big data. Nothing in the training of the typical HIV/AIDS researcher prepares him or her for data analysis on such a scale. Given the data complexity, it is inevitable that major HIV/AIDS research projects will evolve to adopt a multi- disciplinary team science approach. Therefore, there is a need for HIV/AIDS researchers to understand the data science and statistical analysis challenges associated with large data sets, and for statistical, computational and mathematical researchers to understand the challenges specific to HIV/AIDS research. This proposal directly addresses these needs by providing workshops in data science, statistical thinking and multi- parameter data analysis, catered specifically to HIVAIDS researchers, as well as immersive internships in leading HIV/AIDS research laboratories for graduate students in statistics, mathematics, computer science and engineering. Combined, the workshops and internships will build a cadre of HIV/AIDS and quantitative scientists able to effectively bridge disciplinary communication barriers, and form effective and productive collaborative partnerships. If successful, this proposal will serve as a model for the training of the next- generation of inter-disciplinary HIV/AIDS basic and translational science researchers. The proposal fulfills the Courses for Skills Development requirement by offering three modules of workshops every year (equivalent to 36 full days per year) covering data science skills, statistical thinking and multi- parameter data analysis contextualized for HIV/AIDS researchers. The first two modules are targeted at graduate students, postdocs and new faculty in HIV/AIDS research; the third module is for a mixed audience. The proposal fulfills the Research Experiences requirement by providing full-time summer internships for graduate students in the quantitative sciences to conduct research in leading HIV/AIDS laboratories, co- mentored by an HIV/AIDS researcher and quantitative science faculty member. We propose to have 6 such projects each year, each training two interns. Finally, the project fulfills the Mentoring Activities requirement, informally by providing technical expertise, advice and insight from experienced researchers during the extended workshops and internships; and formally by establishing interdisciplinary collaborations and running an annual grantsmanship session for HIV/AIDS biomedical researchers specifically focused on how to write the data analysis plan for an NIH grant submission.
HIV/AIDS research is generating increasingly large and complex data sets. To analyze these data sets, we need the next generation of HIV/AIDS researchers to learn skills in data processing and statistical analysis, as well as increase collaboration with quantitative scientists such as statisticians, mathematicians, computer scientists and engineers. The proposed curriculum will train HIV/AIDS researchers in the data science and statistics skills required to analyze multi-parameter data through a series of hands-on workshops and provide internships to immerse graduate students in the quantitative sciences necessary for research currently underway at leading HIV/AIDS laboratories.
