Using Make for Reproducible and Parallel Neuroimaging Workflow and Quality-Assurance

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Bioinformatics

Background:

• Neuroimaging workflows are complex and require robust management for reproducibility and efficiency.
• Existing pipeline systems can be resource-intensive and demand significant programming expertise.
• Dynamic research environments necessitate adaptable and maintainable workflow solutions.

Purpose of the Study:

• To demonstrate the utility of the Make software tool for programming neuroimaging workflows.
• To highlight how Make can facilitate reproducibility, parallelization, and fault tolerance in neuroimaging.
• To present Make as a more accessible alternative to sophisticated pipeline systems.

Main Methods:

• Utilizing Make, a software development tool, to define neuroimaging workflow construction.
• Employing makefiles with rules to manage dependencies and create directed acyclic graphs.
• Leveraging Make's capabilities for parallel execution and automated updates based on file modifications.

Main Results:

• Make enables the achievement of key pipeline features such as reproducibility, parallelization, and fault tolerance.
• The use of Make significantly reduces the programming demands compared to shell scripting.
• Workflows programmed with Make are easier to debug and maintain in evolving research settings.

Conclusions:

• Make offers a practical and efficient approach to programming robust neuroimaging workflows.
• This method lowers the barrier to entry for creating reproducible and scalable neuroimaging analyses.
• The accompanying manual and open-source project facilitate the adoption of Make in neuroimaging research.