As part of their coursework, students put their DevOps skills into practice by containerizing advanced AI models for protein analysis. Among the tools covered in the OCC (Operating Systems, Code and Computation) course were Placer, which focuses on predicting protein conformational flexibility, and ProstT5, a protein language model capable of establishing relationships between sequences and structures.
These tools, often complex to install and configure, were packaged into Docker containers, ensuring portability and ease of use across different computational environments.
In bioinformatics and computational biology, reproducibility of results is a critical concern. Minor differences in software environments or dependencies can be enough to compromise an analysis. By mastering containerization, students learn how to standardize the execution of complex models and ensure that results can be reliably reproduced, shared, and validated across research settings.
This project clearly demonstrates how DevOps helps bridge the gap between academic research and real-world deployment. By making protein analysis tools more accessible and easier to deploy, students contribute to bringing AI-driven solutions closer to the practical needs of laboratories and research teams, both locally and internationally.
Through projects like this, the Digital Life Sciences major reinforces its interdisciplinary positioning at the crossroads of AI, life sciences, and software engineering. Students develop highly sought-after skills, enabling them to understand scientific challenges, algorithmic models, and the technical constraints associated with deploying advanced digital solutions. The study plan for the major is available here.