What we can learn from deep space communication for reproducible bioimaging and data analysis

Session

Poster Session

Authors

Tatiana Woller (4), Christopher Cawthorne (1), Romain Slootmaekers (2), Ingrid Roig (1), Alexander Botzki (3), Sebastian Munck (3)

Affiliations

1. KU Leuven
2. NSANGA
3. VIB / KU Leuven
4. VIB/KULeuven

Keywords

image analysis, metadata, documentation, space communication, FAIR

Abstract text

Multiple initiatives have attempted to define and recommend the annotation of images with metadata. However, proper documentation of complex and evolving projects is a difficult task, and the variety of storage methods—electronic lab notebooks, metadata servers, repositories, and manuscripts—along with data from different time points of a given project leads to either redundancy in annotation or omissions. 

Therefore, to make modern data management and analysis that is committed to FAIR principles and improved reproducibility a reality, Flanders BioImaging in a team effort with the Flemish supercomputing center (Vlaams Supercomputer Centrum, VSC) and KU Leuven’s central IT infrastructure, is exploring new models for data management, metadata handling, and image analysis. These combined efforts aim to serve the imaging data together with other research data. One result is the development of the ManGO platform running on iRODS, using different metadata schemas and combining metadata across different platforms. 

Here, we discuss how to tackle the problem of redundancy across platforms, taking inspiration from space communication, which uses error-correction protocols based on redundancy for data transmission. We provide a proof of concept using an Artificial Intelligence (AI) language model to digest redundant metadata entries of this manuscript and visualize the differences to complete metadata entries, highlight inconsistencies, and correct human error to improve the documentation for more reproducibility and reusability. 

Consequently, we will reflect on our recent publication on this topic by Woller et al. (Mol. Sys. Biol. 2024; doi: 10.1038/s44320-023-00002-9)

References

Woller T, Cawthorne CJ, Slootmaekers RRA, Roig IB, Botzki A, Munck S.  What we can learn from deep space communication for reproducible bioimaging and data analysis. Mol Syst Biol. 2024 Jan;20(1):1-5. doi: 10.1038/s44320-023-00002-9.