Abstract
Recent advances in sequencing technology enable the capture of gene expression dynamics through longitudinal study designs. However, the field lacks robust analytical tools for these high-dimensional datasets. To address this, we developed bioinformatic methods to determine baseline gene expression variability and identify temporally varying genes (TVGs). Our dynamic cut-off metric enhances detection of differentially expressed genes (DEGs), reducing false positives and negatives, while our TVG scoring system identifies genes with fluctuating expression over time. In a 21-day longitudinal RNA-seq dataset from rats, we identified 502 DEGs and 300 high-confidence TVGs, revealing that repeated blood sampling activates pathways related to bleeding, coagulation, and inflammatory responses. These findings were recapitulated in a 9-day human longitudinal RNA-seq study, which revealed similar pathway enrichments and subject-specific expression dynamics. Additional controls confirmed that these gene expression changes were not induced by handling artifacts, and score threshold analysis enabled a tunable balance between sensitivity and specificity. Our study introduces a robust analytical framework and the largest high-frequency longitudinal RNA-seq dataset of its kind, now publicly available. These resources provide valuable insights into dynamic gene regulation and offer new approaches in systems biology, pharmacogenomics, and translational medicine.