Reducing Theory to Practice: Designing Training Modules and BCO Tools for Adaptation of BioCompute by Novice Audiences Open Access
Downloadable ContentDownload PDF
In recent years, the field of bioinformatics has seen an exponential growth of biological data generation and pipeline development due to the unprecedented scale in the advancement of high-throughput technologies. This increasing growth and complexity of bioinformatics research is accompanied by the urgent need to capture the data generated, analysis mechanisms, procedures used, and various environments of execution. Currently, computational researchers use a variety of bioinformatics software with different environments and parameter settings, which leads to variation in research protocols and difficulty reproducing results. While there is no shortage of open source information on bioinformatics workflows and pipelines, there is a lack of standardized documentation and accessibility. BioCompute is a standardized mechanism of analysis communication, built specifically for collaboration, reproducibility, and recording provenance, and which extends to both data analysis workflows and the development of knowledgebases. BioCompute Objects (BCO) are the artifacts of the BioCompute Framework; they are machine and human readable data objects that adhere to the BioCompute Specification. This curated audit trail of bioinformatics processes is able to support the dynamic elements, complexities and variations intrinsic to bioinformatics research. This level of communication promotes the accessibility and reproducibility of novel and existing pipelines. This transparency and accessibility will further enable interoperability between the diverse groups in academia, industry, and regulatory bodies that exist within the scientific community. To evaluate and further refine the BCO Specification and to make robust for more use cases, training modules and BCO tools can be leveraged. This thesis explores the development of user-friendly platform-dependent and platform-independent BCO creation tools (HIVE BCO App and BCO Editor) and training modules used in the implementation of BioCompute in an academic research environment to generate community engagement data from novice bioinformaticians.