add peak calling assay and protocol

assays/PeakCalling/protocols/PeakCallingProtocol.md

+## Peak calling
+
+Peak calling on predictions and the experimental data was performed with MACS3 (Zhang et al. 2008). The sample bam files of the experimental data per species and dataset were merged. Then peaks were called on the merged bam files with MACS3’s “callpeak” command. The parameters for calling ATAC-seq peaks were the BAMPE format, a q-value of 0.01, keeping all duplicates, using the background lambda as local lambda (“no-lambda”) and the ungapped genome size of the species’ genome assembly (see Supplementary Table S2) as mappable genome size. For ChIP-seq peak calling two parameters, broad and a broad cutoff of 0.1, were added. The chosen q-value was the default 0.05. The ChIP-seq peaks of the species *S.polyrhiza* and *Chlamydomonas reinhardtii* were called using the format BAM instead of BAMPE. MACS3’s “bdgpeakcall” was used to call peaks on the test species predictions in bedGraph file format. The parameters for peak calling were the same MACS3’s “callpeak” determined for the experimental data, i.e. for paired end reads the minimum length and maximum gap are set to the predicted fragment size (Table 5). The cutoff value, threshold of the minimum read coverage to call a peak, was estimated by plotting the average read coverage of predictions around the TSS (see Fig. 5b).
+
+Different cutoff values were also examined. For the ATAC-seq predictions of *A. thaliana*, cutoffs in the range of 1 to 25 with a step of 1 and for *O. sativa* cutoffs in the range of 5 to 200 with a step of 5 and including a cutoff of 1 at the start were chosen. For the ChIP-seq predictions of both species, cutoffs in the range of 5 to 100 with a step of 5 and including a cutoff of 1 at the start were chosen.
+
+The selected parameters of MACS3’s “bdgpeakcall” for each test species and dataset are listed.
\ No newline at end of file