Fig. 1

Baseline sequencing results comparison of fresh (blue) vs. stored (green) rabbit rotator cuff samples at different tear states. A H&E-stained histological sections on the Visium Spatial transcriptomics slide. B Increasing number of spots under tissue decrease mean reads per spot at a given total number of reads. C Absolute sequencing saturation (black, left Y-axis) and relative %-point of saturation added per 1000 reads/spot (gray, right Y-axis), D median genes per spot (black, left Y-axis) and relative median genes added per spot per 1000 reads/spot (gray, right Y-axis) and, E total number of genes detected (black, left Y-axis) and relative number detected per 1000 reads/spot (gray, right Y-axis) suggest that data quality was not affected by 6 years of sample storage. F Unique Molecular Identifier (UMI) and gene counts per sample relative to different clusters/underlying tissue types. The fresh sample was very homogenous and only yielded one high-quality tissue cluster vs. low-quality section border clusters, thus it was excluded from this piece of analysis. An example of such a low-quality section border cluster was included in the 2 weeks sample. Clusters were named based on the underlying tissue type identified using H&E. Different fiber type clusters were named based on differentially expressed MYH isoforms. Data in (F) are medians and 95% confidence intervals. *p < 0.05; **p < 0.01; ***p < 0.001