Dockstore - A Platform for Creating, Sharing, Publishing and Reproducing Computational Science - PROJECT SUMMARY / ABSTRACT The crisis in scientific reproducibility threatens to impede progress in the data-driven fields of genomics and biomedicine, where the scale of data generation creates unprecedented opportunities for discovery but makes tracking complex computational methods difficult. To address this foundational challenge, the Dockstore platform provides a critical integration layer for sharing Findable, Accessible, Interoperable, and Reusable (FAIR) computational methods, unifying container technologies, workflow languages, and source code repositories into a cohesive, searchable index of ready-to-use tools. Having become integral to numerous NIH-funded consortia, including NHGRI’s AnVIL, and having demonstrated substantial growth in content and users, Dockstore is poised to evolve into an even more powerful community resource. This proposal seeks to significantly enhance Dockstore with intelligent features, expanded interoperability, and deeper community engagement to meet the growing demands of modern, data-intensive science. The project is organized into five specific aims. First, we will enhance the developer and workflow authoring experience by deploying AI-assisted tools to semi-automate metadata creation, lowering the barrier to contributing high-quality content, while also expanding support for new content types like Snakemake, RMarkdown, and the linking of AI models. Second, we will deploy AI-powered personalized discovery, curation, and search, implementing a natural language query system for intuitive searching and an intelligent recommendation engine to suggest relevant workflows. Third, we will establish comprehensive metrics and analytics to empower users to make informed decisions, by ingesting and visualizing execution data from partner platforms, including key metrics such as computational cost, runtime, and resource consumption. Fourth, we will implement cross-platform federation and enhanced API support by enabling federated search with international partners like WorkflowHub.eu and strengthening support for interoperability standards such as the GA4GH Tool Registry Service (TRS) API. Fifth, we will strengthen security, compliance, and trust mechanisms, maintaining Dockstore’s FISMA Moderate security posture and implementing new features for workflow security scanning and ORCID integration for identity verification.These efforts will be supported by a comprehensive outreach and dissemination plan focused on improving documentation and training, reaching new audiences through targeted workshops, and gathering user feedback to ensure user-focused development. Collectively, these enhancements will solidify Dockstore’s role as an indispensable hub for reproducible computational science, accelerating discovery and collaboration across the global biomedical research community.
