We study kidney disease progression in female lupus-prone NZB/W F1 mice. Mice were euthanized at 10, 20 or 30 weeks of age and one kidney sample was taken from each, giving 9 samples, 3 per age. The age at euthanasia is in the `group` column and defines three groups of mice: 10w means 10 weeks of age (early stage), 20w means 20 weeks, and 30w means 30 weeks (late disease). There are no wild-type or non-lupus control mice in these files, and no mouse received an experimental treatment.

Each age group is a separate set of mice. Every mouse was sampled once, at the age at which it was euthanized, so the three ages are groups of different animals rather than repeated measurements of the same animals over time, and no mouse contributed more than one sample. The labels rep1 to rep3 in `sample_id` only number the mice within each age; mice with the same number at different ages are different animals with no recorded link. Each mouse gave one library, so there are no technical replicates. No processing or sequencing batch is recorded, and we know of no other variable that matters; the remaining columns hold one value for all samples.

The data are bulk poly(A) mRNA-seq of kidney tissue, given as raw gene-level read counts from HTSeq-count after STAR alignment to mm10; the five HTSeq summary rows at the end of the original table have been removed.

Our question is which genes differ in kidney expression in the 30-week-old mice (30w, late disease) compared with the 10-week-old mice (10w, early stage). We are asking about the difference between these two ages, not about an overall trend across all three. All nine samples are part of the analysis, including the 20w mice.