| Results | One potential explanation for why the SVD and NMF components tracked cell-state proportions is that the components were identifying genes differentially expressed between cell states. |
| Results | In contrast, SVD component 2 identified the only two genes that were strongly differentially expressed between states B and C (HOXA5, FOXOI; Fig 2E); these two genes, HOXA5 and FOXOI, were respectively down and up in state B relative to state C, and were expressed near median levels in state A. |
| Results | Thus, the highest loadings of SVDl in this idealized experiment marked genes differentially expressed between luminal and basal cells, including the established luminal markers GATA3 and STAT5A. |
| Abstract | Our computational statistical method is well suited to meta-analyses as there is no requirement for transcripts to pass thresholds for significant differential expression between time points, and it is agnostic to the number of time points per dataset. |
| Discussion | The model definition and selection methodology we present is not limited by eXpression level, for example by tests for differential eXpression , nor do we rely upon the arbitrary thresholding that is common in clustering analyses. |
| Introduction | This class of transcripts is currently understudied, but lncRNAs are differentially expressed during differentiation, are preferentially localised in chromatin and have been proposed to ‘f1ne-tune’ cell fate via their roles in transcriptional regulation [14—16]. |