SpaceSenAI

Inspiration

Space radiation poses a challenge to human health during long-duration spaceflight. Exposure to ionizing radiation can cause molecular damage and may contribute to cellular senescence, a state in which cells stop dividing and undergo changes that can affect surrounding tissues.

We developed SpaceSenAI to explore how publicly available NASA transcriptomic data and interpretable artificial intelligence could help researchers investigate radiation-associated senescence and identify molecular targets for further study.

What it does

SpaceSenAI is a computational research project designed to connect space-radiation biology with data-driven molecular analysis.

The proposed platform aims to:

• Analyse gene-expression data from NASA's Open Science Data Repository.
• Identify genes and biological pathways associated with radiation exposure and cellular senescence.
• Apply interpretable computational methods to prioritise candidate molecular targets.
• Present the findings through an accessible, research-oriented dashboard.

The goal is to support hypothesis generation and research prioritisation, not to provide clinical diagnoses or treatment recommendations.

How we built it

The project is based on publicly available NASA transcriptomic datasets, including OSD-993 (bulk RNA sequencing) and OSD-992 (single-cell RNA sequencing).

Our planned analytical workflow combines data quality assessment, gene-expression analysis, pathway enrichment, and computational prioritisation. Where appropriate, findings from bulk data can be explored alongside single-cell evidence.

Python and open-source bioinformatics and machine-learning tools provide the foundation for the analytical pipeline. The dashboard is intended to make the results easier to explore and interpret.

Challenges

A major challenge is working with high-dimensional biological data, where thousands of genes may vary across a relatively small number of samples. This creates risks of overfitting and makes careful statistical analysis essential.

Other challenges include distinguishing radiation-associated signals from experimental and dataset-specific effects, interpreting candidate genes biologically, and avoiding unsupported conclusions from computational predictions.

Our approach is to prioritise transparent analysis, pathway-level interpretation, and appropriate validation rather than treating computational predictions as established biological findings.

What we learned

This project has helped us connect space biology, transcriptomics, bioinformatics, and machine learning. It has also highlighted the importance of data quality, reproducibility, biological interpretation, and clearly communicating the limitations of computational research.

Future scope

Future development will focus on completing and validating the analysis pipeline, improving the interactive dashboard, and investigating whether prioritised molecular signatures are reproducible across independent datasets.

SpaceSenAI aims to provide a starting point for further research into radiation-induced cellular senescence and potential countermeasure strategies.

Built With

• bioinformatics
• data
• data-science
• machine-learning
• nasa-open-science-data-repository
• numpy
• pandas
• python
• scikit-learn
• transcriptomics