Analyzing the relationship between gene expression and phenotype in space-flown mice using a causal inference machine...

Area of Science:

• Space biology
• Genomics
• Machine learning

Background:

• Spaceflight induces detrimental health effects, including liver dysfunction in rodents.
• Transcriptomic changes are observed alongside liver dysfunction, but correlation-based analyses may yield spurious results.
• Existing tools lack the ability to distinguish true correlations from false ones in transcriptomic data.

Purpose of the Study:

• To apply a novel machine learning ensemble of causal inference methods to identify robustly correlated genes with liver dysfunction phenotypes.
• To uncover novel molecular targets and genes associated with spaceflight-induced liver dysfunction.
• To improve upon traditional differential gene expression analyses.

Main Methods:

• Utilized the Causal Research and Inference Search Platform (CRISP), a machine learning ensemble for causal inference.
• Applied CRISP to high-dimensional transcriptomic and histological data from space-flown rodents.
• Focused on identifying genes robustly correlated with a lipid density phenotype.

Main Results:

• CRISP identified key genes and molecular targets not previously detected by traditional analyses.
• The identified genes are strongly correlated with the lipid density phenotype in space-flown rodents.
• This causal inference approach provides a more accurate understanding of gene-phenotype relationships.

Conclusions:

• Novel genes and molecular targets involved in spaceflight-induced liver dysfunction were identified.
• The findings offer potential new targets for developing countermeasures against adverse health effects of space travel.
• This study highlights the utility of causal inference methods in analyzing complex biological data from space exploration.