Driving Innovation in Research Software Engineering:
The Revamped NeuralActivityCubic Tool
At Indoc Research, our mission is to support cutting-edge scientific research by providing high-quality, custom software solutions. Recognized by governing bodies in Canada and Germany as a not-for-profit organization (German: gemeinnützig), we are eligible for collaboration on most research grants. As part of our Research Software Engineering (RSE) services, we collaborate with researchers to develop, maintain, and optimize software tools that enhance scientific discovery. Here, we present our recent collaboration on the NeuralActivityCubic (NA³) tool, which exemplifies how we bridge the gap between science and technology to meet the evolving needs of the scientific community.
Addressing the Challenges of Calcium Imaging Analysis
Calcium imaging is a widely used method in neuroscience to assess the effects of pharmacological agents on neuronal excitability and activity. However, analyzing calcium imaging data is complex, particularly in high-throughput screenings, where large datasets must be processed quickly and automatically. Low signal-to-noise ratios further complicate the identification of true neuronal activity.
Originally developed by Prada et al. at the University of Würzburg, NeuralActivityCubic (NA³) was designed to tackle these challenges. The tool excels at detecting subtle calcium signals, especially in regions like neurites, where signals are often near the noise floor. Despite its value, NA³ became unusable due to accumulated software dependency issues, prompting the need for a complete overhaul.
Revamping NA³: Enhancing Usability, Performance, and Accessibility
In collaboration with the original NA³ developers, we aimed to make the tool more accessible, maintainable, and adaptable for modern analysis demands. Our collaborative approach is summarized in the following video:
Key improvements include:
• Migration to Python: We ported NA³ from R and Bio7 to Python, addressing the dependency issues caused by outdated R packages. Python’s growing prominence in life sciences ensures that NA³ will remain a relevant and powerful tool in the years to come.
• Optimized Performance: By incorporating multi-threading and batch processing, we achieved a significant increase in computation speed, essential for high-throughput datasets.
• Design Patterns for Maintainability: To future-proof the tool, we used design patterns like Model-View-Controller (MVC) and the Factory Pattern. These ensure low coupling and high cohesion, making the software easier to maintain, extend, and share.
• Cloud Accessibility: We containerized the tool for cloud deployment, allowing users to avoid local installation and enabling remote access to computing resources.
Curious about the revamped version of NA³? Please visit our GitHub repository or try it for free on Binder! Our poster presented at the 37th ECNP is available for download here.
Proven Impact: Applying NA³ to Pharmacological Screenings
The updated NA³ was validated through a collaboration with researchers at the University Hospital of Würzburg, where calcium imaging was performed on sensory neurons derived from dorsal root ganglia of adult mice. The neurons were exposed to various pharmacological agents, including the ApoA1 mimetic peptide D-4F, to evaluate their effects on neuronal activity and excitability. Our findings are summarized in the following short video:
Key findings include:
• High Sensitivity: NA³ successfully detected both action potential-induced calcium signals and more subtle, non-spike-like events, even at low signal-to-noise ratios.
• Enhanced Performance: The revamped NA³ tool significantly reduced analysis time, making it suitable for high-throughput screenings.
• Pharmacological Insights: Our analyses indicated that D-4F increased both neuronal activity and excitability, as measured by events per minute and the calculated variance area.
These findings underscore NA³’s potential in evaluating how pharmacological agents influence neuronal behavior, making it an essential tool for neuroscience research. Try NA³ yourself for free via Binder – no installation required! Our poster presented at the 37th ECNP is available for download here.
Concluding Thoughts: A Model for Future Research Software Development
The NA³ project highlights the importance of long-term software maintenance in scientific research. Many tools, like NA³, are developed as secondary outputs of research grants, but their utility is often limited by a lack of ongoing funding for updates. At Indoc Research, we fill this gap by providing the expertise needed to ensure tools like NA³ remain accessible and performant.
Our collaboration with the NA³ team demonstrates how combining scientific expertise with professional software engineering leads to more robust, maintainable, and impactful research tools. This approach serves as a model for ensuring that research software remains sustainable and powerful.
Learn More and Collaborate with Us
Revamping an existing tool, like we did for NeuralActivityCubic, is just one example of how we at Indoc Research can support you in championing Research Software Engineering. Whether you're interested in learning more about NA³ or looking to collaborate on developing software for your research, we’d love to hear from you! Please contact us to share your ideas and questions or visit us on GitHub.