The recount3 Python package for programmatic access to uniformly processed RNA-seq data

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• The recount3 resource provides extensive uniformly processed RNA-seq data.
• Existing access methods primarily cater to the R/Bioconductor ecosystem.
• There is a growing need for efficient Python-based tools for transcriptomic data analysis.

Purpose of the Study:

• To introduce the recount3 Python package for accessing and utilizing recount3 data.
• To provide Python users with native, efficient tools for RNA-seq data discovery and download.
• To bridge the gap between large-scale transcriptomic datasets and Python-based machine learning pipelines.

Main Methods:

• Development of a Python package with an application programming interface (API) and command-line interface (CLI).
• Implementation of uniform resource locator (URL) resolution and persistent on-disk caching.
• Automatic parsing of downloaded data into analysis-ready formats (Pandas DataFrames, BiocPy RangedSummarizedExperiment objects).

Main Results:

• The recount3 Python package facilitates discovery, download, and materialization of recount3 resources.
• The software handles URL resolution, caching, and data parsing efficiently.
• Analysis-ready data structures are automatically generated for Python users.

Conclusions:

• The recount3 Python package significantly lowers the barrier for large-scale RNA-seq data utilization in Python.
• It enables seamless integration of public transcriptomic data into Python-based computational workflows.
• This facilitates the application of modern machine learning techniques to vast RNA-seq datasets.