Foreword
Study goals and section’s structure
1 Goal of this study
The main goal of this study was to explore the effect of Intermittent Hypoxia (encoded by the IH level of the Condition variable) on the expression of certain genes of interest, at various developmental stages, and in multiple layers of the cerebellum. We also correlate these changes in gene expression with cell death mechanisms.
2 General methodology
This project included RT-qPCR and immunohistochemistry data, which were analysed in R (R Core Team, 2023).
Data were modeled through the Generalized Linear Mixed Model (GLMM) framework, using the glmmTMB package (Brooks et al., 2017). Random intercepts were added to account for the correlation between pseudo-replicates, or intra-class correlation stemming from plate/experiment effects.
Model diagnostics were done using the DHARMa (Hartig, 2022) & performance (Lüdecke et al., 2021) packages, and estimated marginal means/contrasts were computed with the emmeans package (Lenth, 2022).
3 Structure of this section
Similarly to the Data section, the Analysis section is sub-divided into two sub-sections, for PCR & IHC analyses.
3.1 PCR
Here, we analyzed the changes in expression of two panels of genes: a Neurodevelopment (ND) panel and an Oxidative Stress (OS) panel.
Refer to the relevant sub-section of the Data section to see which genes and pathways are studied in each panel.
Within each panel, we define a simple linear regression model to test the effect of Condition on a gene’s expression, which is operationalized through the DCq variable, representing the difference between the quantification cycles of the gene of interest and a combination of relevant housekeeping genes. Random effects were added when we suspected potential clustering effects, such as the possible correlation or hierarchical dependencies stemming from the experimental design.
That model was then applied to each Gene of a given Stage and Layer.
Each gene panel will be divided in:
Model fitting & diagnostics: which model was fit to the data of this panel’s genes, and how well does the model fit the observed data.Model analysis: the models’ coefficients and contrasts of interests. This part several plots showing which genes are up or down-regulated genes, as well as the fold-change timelines showcased in the article.
3.2 IHC
Here, we analyzed the apoptotic marker cleaved Caspase-3 to further explore cell death mechanisms induced by Intermittent Hypoxia. Meanwhile, the marker Calbindin allowed us to monitor the effect of IH on Purkinje Cells.
Both Capsase and Calbindin sub-sections contain multiple responses of interest, which were modeled independently.
Each response of interest will be divided in:
Model fitting & diagnostics: which model was fit to the data of this variable, and how well does the model fit the observed data.Model analysis: the models’ coefficients, marginal means, and contrasts of interests. This part includes the boxplots showcased in the article.