Simple Model: ~condition

Model Configuration

The simplest differential expression model compares gene expression between levels of a single factor, such as Treatment versus Control.

When to Use

Your experiment has one main factor of interest (e.g., Treated vs. Control, Knockout vs. Wild-type).
You want to compare two or more groups without adjusting for any additional covariates or batch effects.
This is your first model and you want to verify the pipeline works end-to-end before adding complexity.

Main factor: the metadata column containing group labels (e.g., "Treatment").
Reference level: the baseline group that serves as the denominator of the fold-change calculation (e.g., "Control", "DMSO", "WT").
Test level: the comparison group (e.g., "Treated", "Doxorubicin", "KO").
Design preview is displayed in the configuration dialog (e.g., ~condition).

easyCris runs size-factor normalisation, gene-wise dispersion estimation, and shrinkage fitting as part of the analysis pipeline.
The results table contains: gene name, baseMean, log2FoldChange, lfcSE, pvalue, and padj.
A positive log2FoldChange means the gene is expressed more highly in the test group; a negative value means it is higher in the reference group.
The padj column (Benjamini-Hochberg adjusted p-value) is the primary metric for calling statistical significance.

Swapping the reference and test levels flips the sign of log2FoldChange but does not change significance. Choose the biological baseline as the reference.
Fewer than three replicates per group severely limits statistical power. Aim for at least three biological replicates.
The reference level should always be the untreated, wild-type, or control condition so that positive fold changes represent the biological effect of interest.

Zhu, A., Ibrahim, J. G., & Love, M. I. (2019). Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences. Bioinformatics, 35(12), 2084-2092. doi:10.1093/bioinformatics/bty895.