Plugin Architecture
Stages are loaded dynamically at runtime. Add, remove, or substitute a step without touching the core engine.
Bioinformatics Research Group · FIU
A lightweight,
plugin-based pipeline
for microbiome analysis.
PluMA lets researchers implement an algorithm in their programming language of choice and test it within a larger pipeline — alongside stages written in other languages that use different file formats. The core is infinitely extensible through dynamically loaded plugins.
Abstract
What is PluMA?
PluMA — Plugin-Based Microbiome Analysis — is a pipeline framework from the BioRG group at Florida International University. It began as a microbiome-analysis tool and has grown into a general-purpose environment for constructing lightweight, composable, reproducible bioinformatics pipelines.
A PluMA pipeline is a sequence of plugins, each a self-contained stage authored in a language of the researcher's choosing. The core engine discovers and loads plugins on the fly, so stages in C++, Python, Perl, R, or Java can be freely combined within a single workflow. New algorithms can be dropped in, tested, and benchmarked without modifying the core — and plugins developed by others can be reused across projects without porting.
Design Principles
Why researchers choose PluMA
Language-Agnostic
Implement plugins in C++, Python, Perl, R, or Java. Stages in different languages interoperate through a shared file interface.
Lightweight
The core executable is under 200KB. Download only the plugins your pipeline needs from the pool of over 1,500.
CPU & GPU Support
Dedicated GPU plugins accelerate computationally intensive stages without changing the pipeline definition.
Composable Pipelines
Pipelines are declarative — stages execute sequentially and can be reordered or swapped through configuration alone.
Research-Proven
Used across metagenomics, metatranscriptomics, metabolomics, multiomics, and epigenetics studies.
Citation