National Technical University of Athens (NTUA)
Μελέτη της επίδρασης του βαρυτικού πεδίου στη γονιδιακή έκφραση και ρύθμιση με τη χρήση υπολογιστικών προτύπων και μεθόδων μηχανικής μάθησης
Abstract
dc:descriptionThe upcoming human deep-space exploration missions necessitate an in-depth understanding of the effects of spaceflight conditions on human physiological systems. The complexity of organisms is directly linked to the complexity of the environment that sustains them. It appears that gravity not only curves the space-time continuum but also the biological continuum.The NASA GeneLab Open Science Repository (OSDR) (https://osdr.nasa.gov/bio/) serves as a valuable resource, hosting data derived from model organisms and human experiments conducted during spaceflights and terrestrial microgravity analogs. Machine Learning applications have the potential to maximize the optimal utilization of existing datasets to understand and ultimately mitigate physiological abnormalities during long-duration missions. In this study, we identified enriched terms and pathways associated with significantly differentially expressed genes in each organism and across orthologous genes. Additionally, we constructed AI-ready merged meta-datasets consisting of musculoskeletal tissues from Mus musculus and Homo sapiens, while simultaneously applying multiple Supervised Learning models to classify significantly overexpressed and underexpressed genes. Furthermore, we explored Transfer Learning by pre-training a model on the Mus musculus dataset and subsequently fine-tuning it on the Homo sapiens dataset. This doctoral dissertation highlights the significant biological impacts of altered gravity conditions on gene expression and the dysregulation of biological pathways. The findings in Mus musculus indicate disruptions in cell cycle regulation, associations between dysregulated signaling pathways and cancer development, as well as an increased risk of chronic inflammation and diabetic complications.Correspondingly, in Homo sapiens, the results point to systemic metabolic disorders, cellular aging, diabetes, impaired tissue repair mechanisms, and cardiovascular vulnerabilities, such as ventricular tachycardia and long QT syndrome. The comparative meta-analysis of orthologous genes between species reveals conserved pathways affecting muscle and skeletal integrity, metabolic homeostasis, inflammatory responses, kidney and heart function. These biological impacts emphasize the need for targeted countermeasures to mitigate the systemic and long-term health risks for organisms experiencing microgravity conditions during extended spaceflights. Finally, this study showcases the potential of leveraging Machine Learning and Transfer Learning methodologies to address the challenges of analyzing specialized datasets in space research. The ultimate goal is to bridge the gap between species, facilitating knowledge transfer from model organisms to humans, and overcoming the limitations imposed by the scarce availability of human data.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Adamopoulos, Konstantinos
- Αδαμόπουλος, Κωνσταντίνος
Subjects
dc:subject × 19- Διαστημική πτήση
- Μικροβαρύτητα
- Διαστημική Βιολογία
- Μικροσυστοιχίες
- Διαφορική γονιδιακή έκφραση
- Μυοσκελετικό σύστημα
- Μηχανική Mάθηση
- Spaceflight
- Microgravity
- Space Biology
- Differential gene expression
- Microarrays
- Machine learning
- Musculoskeletal systems
- Βιοϊατρική Μηχανική
- Επιστήμες Μηχανικού και Τεχνολογία
- Medical Engineering
- Engineering and Technology
- Biomedical Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/59170