This project analyzes NASA GeneLab RNA-seq transcriptomics datasets using Python, bioinformatics workflows, data visualization, and machine learning techniques.
The analysis focuses on gene expression patterns observed in biological samples exposed to spaceflight-related conditions.
By combining transcriptomics, exploratory data analysis, and machine learning, this project investigates how biological systems may respond to environmental stressors associated with spaceflight.
✔ NASA GeneLab RNA-seq analysis
✔ PCA transcriptomic exploration
✔ Gene expression profiling
✔ Heatmap visualization
✔ Comparative transcriptomics
✔ Machine learning clustering
✔ Reproducible Python workflow
✔ Space biology data analysis
Principal Component Analysis (PCA) revealed substantial variability between biological samples exposed to spaceflight-related conditions.
Several genes showed consistently elevated expression levels across analyzed datasets and contributed strongly to overall expression patterns.
K-Means clustering grouped genes into distinct expression profiles, suggesting underlying biological programs and expression states.
Gene expression analysis provides insight into molecular adaptations potentially associated with microgravity, radiation exposure, and physiological stress.
Gene expression distributions across RNA-seq samples.
PCA visualization of transcriptomic variability across samples.
Heatmap showing expression patterns of highly expressed genes.
Most highly expressed genes identified within the OSD-245 dataset.
Comparison of highly expressed genes across analyzed spaceflight datasets.
NASA GeneLab RNA-seq datasets:
- GLDS-168
- GLDS-245
Sources:
https://osdr.nasa.gov/bio/repo/data/studies/OSD-168
https://osdr.nasa.gov/bio/repo/data/studies/OSD-245
These datasets contain transcriptomic measurements from mouse biological samples exposed to spaceflight-related environments.
- RNA-seq transcriptomics
- Gene expression analysis
- Comparative transcriptomics
- Principal Component Analysis (PCA)
- Machine Learning
- Clustering
- Exploratory Data Analysis (EDA)
- Python
- Pandas
- NumPy
- Matplotlib
- Seaborn
- Scikit-learn
- Jupyter Notebook
nasa-spaceflight-ai/
│
├── figures/
│ ├── osd245_expression_profiles.png
│ ├── pca_spaceflight.png
│ ├── top_genes_heatmap.png
│ ├── top10_genes_osd245.png
│ └── top10_spaceflight_genes.png
│
├── nasa_spaceflight_ai.ipynb
├── nasa_comparative_transcriptomics.ipynb
├── quick_analysis.py
├── README.md
RNA-seq transcriptomics datasets were loaded into Pandas DataFrames and prepared for downstream analysis.
Gene expression distributions were explored to identify highly expressed genes and transcriptomic trends.
PCA was applied to reduce dimensionality and visualize sample-to-sample variability.
Heatmaps were generated to visualize expression patterns and identify dominant transcriptional programs.
K-Means clustering was used to identify transcriptomic groups and expression patterns within the data.
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler
X = df[['TPM', 'FPKM', 'expected_count']]
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)
kmeans = KMeans(n_clusters=3, random_state=42)
df['cluster'] = kmeans.fit_predict(X_scaled)The RNA-seq analysis revealed substantial variability in gene expression profiles across NASA spaceflight-related biological samples.
Principal Component Analysis identified transcriptomic differences between samples, suggesting potential biological responses to environmental stressors associated with spaceflight.
Highly expressed genes may be involved in:
- cellular stress responses
- metabolic adaptation
- mitochondrial function
- immune regulation
- tissue remodeling
- radiation response pathways
The clustering analysis further demonstrated that transcriptomic profiles can be grouped into distinct expression states, highlighting the usefulness of machine learning approaches for exploring complex biological datasets.
- astronaut health monitoring research
- biological stress response analysis
- radiation response studies
- molecular adaptation research
- biomarker discovery
- disease transcriptomics
- cancer genomics research
- personalized medicine research
- transcriptomic pattern discovery
- machine learning for genomics
- biological data exploration
- large-scale omics analysis
- RNA-seq analysis
- Transcriptomics
- Gene expression profiling
- Biological interpretation
- Principal Component Analysis (PCA)
- Machine learning clustering
- Data visualization
- Exploratory data analysis
- Python
- Pandas
- NumPy
- Scikit-learn
- Matplotlib
- Seaborn
- Jupyter Notebook
- Differential gene expression analysis
- Pathway enrichment analysis
- Gene set enrichment analysis (GSEA)
- Deep learning approaches for transcriptomics
- Multi-omics integration
- Interactive dashboards
- Explainable AI methods for biological interpretation
Clone the repository:
git clone https://github.com/ag48665/nasa-spaceflight-ai.gitInstall dependencies:
pip install -r requirements.txtLaunch Jupyter Notebook:
jupyter notebookThis repository is provided for educational and portfolio purposes.
Agata Gabara
MSc Bioinformatics Student
Research Interests:
- Transcriptomics
- Space Biology
- Cancer Genomics
- Computational Biology
- Machine Learning for Life Sciences
GitHub: https://github.com/ag48665
LinkedIn: https://www.linkedin.com/in/agatha-gabara-06494a37/