We differentiated the human iPSC line LON71 toward thymic epithelial progenitors (TEPs) with an iPSC-to-TEP protocol and harvested cells on days 0, 1, 2, 3, 7, 9, 10, 14, 15, 17 and 22 (1 to 6 samples per day, 29 in all). The harvest day is in the `time` column (the `timepoint` column repeats it); day 0 is the undifferentiated iPSC stage before induction, and the other days count from induction. Four primary pediatric thymic epithelial cell (TEC) samples were added for comparison; their `time` is empty. The `group` column only separates iPSCs (iPSc), differentiating cells (iPSc-to-TEP) and TECs (Thymus).

Each LON71 sample is one harvest of cultures on one day, with its own RNA sample and library, and no technical replicates are recorded. All LON71 samples come from this one cell line, but no column identifies a differentiation or culture, and as far as we know no two samples come from the same culture.

GEO gives every sample a batch number (`batch`, repeated in `src_batch`: 1, 2 or 3) without saying what it stands for. All samples of days 0, 3 and 10 have batch 2, so there is no batch variable among the samples of this question, and we have no other record of processing batches. We have no other variable we consider relevant either; the remaining columns are GEO annotations. The data are bulk 3′ RNA-seq (3′ DGE-seq), given as raw UMI counts per gene (RefSeq genes); two batch 3 samples, from days 14 and 17, have no counts for 3,827 genes.

For this question the harvest days are the groups we compare, not a time trend, so there is no separate time variable. We want to find the genes that are differentially expressed at day 10 of the differentiation compared with day 0, the undifferentiated iPSCs, using the samples of days 0, 3 and 10 ("day 0", "day 3" and "day 10" in `time`: 3, 4 and 4 samples). The other days and the TEC samples are not part of this question.