Meta-analysis of transcriptomic datasets using benchmark dose modeling shows value in supporting radiation risk...

Area of Science:

• Radiation biology
• Toxicogenomics
• Bioinformatics

Background:

• Benchmark dose (BMD) modeling quantifies biological effects from stressors.
• Transcriptional data analysis using BMD tools can derive BMDs for genes and pathways.
• Previous work demonstrated BMD modeling's utility in radiation research with internal datasets.

Purpose of the Study:

• To assess the reproducibility of BMD values for genes and pathways across different ionizing radiation exposure scenarios.
• To compare BMD values derived from transcriptomic data with those from cytogenetic endpoints.
• To explore correlations between BMD outputs and radiation parameters like dose-rate, quality, and cell type.

Main Methods:

• Retrieved and compiled transcriptomic studies related to ionizing radiation from the Gene Expression Omnibus (GEO).
• Filtered and analyzed datasets using BMDExpress software.
• Employed graphic visualization techniques to correlate BMD outputs with experimental parameters.

Main Results:

• Identified common genes and pathways with low and high dose thresholds across diverse studies.
• Higher BMD values were linked to immune response and cell death pathways; lower values to DNA damage response (e.g., TP53 signaling).
• Transcriptional BMD values from in vivo and in vitro datasets aligned with cytogenetic endpoint BMDs, especially within similar dose ranges.

Conclusions:

• BMD modeling provides consistent and meaningful outputs across different experimental models for radiation research.
• Reproducibility is high for genes and pathways, particularly those below the 25th percentile of dose distribution.
• The methodology supports robust dose-response assessment in radiation toxicology.