CyberTraining: Pilot: Cloud-Enabled Computational Meta-Research Training for Biomedical Evidence Evaluation

The reliability of biomedical evidence is fundamental to the progress of science and the integrity of public health research. Problems such as unpublished results, switched outcomes, and unreported harms are widespread and well documented, but current training methods cannot address them systematically because they rely on time-intensive manual review. This project supports the national interest by building a new generation of biomedical researchers capable of using advanced computing tools to assess, audit, and improve the quality of biomedical research evidence. The project intentionally recruits students from EPSCoR states, rural institutions, and non-research-intensive universities. By delivering all training through cloud-accessible infrastructure, the program gives students at under-resourced institutions the same computational capabilities as those at well-funded research universities. All curriculum materials, software tools, and evaluation data will be openly shared so that other institutions can adopt and adapt this training model at no cost, extending the benefit of the investment far beyond the students directly trained. This CyberTraining Pilot project develops and rigorously tests a cloud-enabled, computationally rigorous training model for biomedical and public health learners at Oklahoma State University Center for Health Sciences. Over 24 months, the project will train two summer cohorts of approximately 15 students each in an intensive 8-week program that integrates cyberinfrastructure skills with applied meta-research content. Trainees will develop practical competencies in reproducible workflows using Docker, Git and GitHub, and Jupyter and RStudio; distributed data access through public APIs including ClinicalTrials.gov and PubMed; and automated validator pipelines for auditing clinical trial transparency and reporting quality. Cloud computing resources replace the need for local high-performance computing infrastructure, enabling equitable participation from geographically distributed institutions. The project will assess whether trainees acquire measurable cyberinfrastructure competencies using validated pre- and post-training instruments, evaluate the feasibility and equity of cloud-based delivery, and examine whether automated validation tools reliably complement human review for evidence auditing. An independent evaluation will generate evidence on effectiveness, scalability, and cost that will inform a subsequent implementation-level expansion. All open educational resources, validator tools, and reproducible workflow templates will be deposited in public repositories for community adoption. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.