June 14th, 2019
Topic: BioCompute: A Standardized Method to Communicate Bioinformatic Workflow Information and Ease Organizational Burden
Presenter: Jonathan Keeney, MD, Executive Steering Committee for BioCompute, GWU
Abstract:
The rapid explosion of the availability of data and analysis tools has driven remarkable insights in the life sciences and beyond, but the rapidly evolving vocabulary of data analysis has also catalyzed unpredictability in communicating analysis pipelines, which heavily contributes to the inability to reproduce results in this space. Current methods of communicating methods in bioinformatics are ad hoc and may be clear to the original architect of the pipeline, but are often very unclear to anyone else, and may be extremely difficult to reproduce. These sorts of inefficiencies can lead to extraordinary burden in understanding or reproducing a bioinformatic pipeline, adding unnecessary time and money, and potentially introducing uncertainties into the way that patient data is handled. BioCompute is a community effort to produce a stable, predictable standard that will eliminate communication inefficiencies and reduce the operational burden of interpreting or reproducing a pipeline. BioCompute is a human- and machine-readable standard that can be written into any analysis platform. A standardized mechanism for understanding bioinformatics workflows has other benefits, including compatibility with the private sector by allowing workflows to be licensed and distributed without exposing the details in the workflow. This talk will introduce a theoretical description of BioCompute, describe examples, and talk about the current state of standardization efforts.