Skip to main content
. 2018 Mar 1;19:169. doi: 10.1186/s12864-018-4559-3

Fig. 1.

Fig. 1

Diagnostic plots for four representative ATAC-seq datasets: SRR891270, SRR3295017, SRR5720369 and SRR580802. (a, d, g and j) Size distributions of sequenced fragments with reads passed filtering criteria for each library. (b, e, h and k) Heatmaps showing the distributions of signals around transcription start sites (TSSs), resulting from inferred nucleosome-free fragments and nucleosome-bound (mono-, di- and tri-nucleosome) fragments. To plot TSS-associated signals arising from nucleosome-bound fragments, fragments associated with di- and tri-nucleosomes were split into two and three sub-reads in silico, respectively. (c, f, i and l) Smoothed histograms of signals showing in b, e, h and k. The sample corresponding to SRR891270 was optimally transposed by Tn5, preloaded with sequencing adapters, while the sample resulting in SRR580802 was over-transposed. The other two datasets were resulted from sub-optimal transposition. Biased size selection could have occurred during library preparation for SRR5720369. Shown here are signals around TSSs on the human chromosomes 1 and 2