From: MCMSeq: Bayesian hierarchical modeling of clustered and repeated measures RNA sequencing experiments
Number of datasets | 10 |
Number of features per dataset | 15,000 |
Sample sizes | 3, 5, and 10 per group |
 | (6, 10 and 20 total) |
Model parameters | Â |
βg1: Difference in log(expression) between | 0 (all features) |
treatment and control at baseline | Â |
βg2: Change in log(expression) over time | 0 (all features) |
in the control group | Â |
βg3: Difference in change in log(expression) | 0 (80% of features) |
over time between the treatment and | ES1 (10% of features), -ES (10% of features) |
control groups | Â |
βg0, αg, \(\sigma _{gb}^{2}\) | Drawn from an empirical distribution based on human |
 | samples in real RNA-Seq data sets with repeated measures |
Additional comparisons | Â |
βg1+βg3: difference in log(expression) | 0 (80% of features) |
between control and treatment at | ES (10% of features), -ES (10% of features) |
follow-up | Â |
βg2+βg3: Change in log(expression) | 0 (80% of features) |
over time in the treatment group | ES (10% of features), -ES (10% of features) |