Researchers can engage the U-BRITE Analysis Gateway via a number of different abstractions we describe below.
The Jupyter Notebook is “an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text. Uses include: data cleaning and transformation, numerical simulation, statistical modeling, data visualization, machine learning, and much more“.
We recommend that research teams consider doing their scripting in these “data science notebooks”, an abstraction that may improve shareability and reproducibility of data analyses and research results.
- Access Jupyter Notebook via Open On-Demand in your web browser at https://rc.uab.edu.
Reusable Pipelines Directory
Many bioinformatics practitioners and their respective research teams here at UAB have been hard at work developing, documenting, and publishing data analysis pipelines that can be co-developed with, or shared with, and re-used by other teams.
The Reusable Pipelines Directory attempts to catalog and link to pipelines of interest to the U-BRITE target audience.
- Explore the QWRAP microbiome data analysis pipeline from CCTS Informatics.
- Read the QWRAP v3 paper.
- Access QWRAP source code at https://gitlab.rc.uab.edu/CCTS-Informatics-Pipelines/QWRAP.
- Browse other pipelines and templates from CCTS Informatics at https://gitlab.rc.uab.edu/CCTS-Informatics-Pipelines.
- Explore the AGHIpop GWAS pipeline for analyzing data from the Alabama Genomic Health Initiative.
- Explore single cell RNAseq data analysis tutorial from Dr. Min Gao in the Informatics Institute.
- View workshop slides, run live Jupyter Notebook, and access source code at https://dx.doi.org/10.5281/zenodo.3700066.
- Browse contributed Hackathon projects, example notebooks, technology demos at https://gitlab.rc.uab.edu/U-BRITE.
- Requires U-BRITE user access.
Smart API Directory
As data analysis pipelines are developed, it is often desirable to make use of existing libraries of code rather than writing new code that may not be as well-tested or documented. As more of these libraries are made available as modern web services, authors of new pipelines can call them over the network via a web service interface (often a REST API), even if those services are implemented in a different programming language than that of the pipeline. This approach promotes further re-use of existing code and can greatly expand the number of “building blocks” available to the pipeline author.
The Smart API Directory attempts to catalog and link to APIs (Application Programming Interfaces) of interest, including those written by the U-BRITE team.