Latencies were only measured for significant peaks.
The frequency of ACGT (N) ACGT was assessed for significant peaks by box and whiskers plots, with 10%and90%0% whiskers.
Additionally, for significant peaks in each time window, we extracted the voxel intensity at the source peak as the log-transformed squared standardized magnitude of the estimated current density.
We identified a small number of significant peaks for FUS (52 peaks) and EWS (133 peaks).
Thus, a score alignment threshold of 30 in the BLAST search and a cut-off threshold of 70 in alignments accumulation were used as standard parameters for the selection of significant peaks identifying putative coding sequences.
The results comprise lists of significant peaks and reports for motif enrichment.
The E-value for estimating significant peaks was set to 100.
A threshold of q < 0.1 was set for calling significant peaks, which were annotated within the nearest 50 kb upstream of genes (including the gene itself) using PeakAnalyzer (Salmon-Divon et al., 2010).
For example, a region on chromosome Z at positions 12,660,165, 15,218,811, 16,201,782, and 23,950,413 (| | ≥ 4.74, p ≤ 1.04.10-6) accounted for 4 significant peaks only in females, while the other remaining 6 significant positions were found on chromosomes 1 and 2, and within unassigned scaffolds.
We use the same notation π to denote the binding probability of probes in the ChIP data, for which significant peaks are to be called, where the π used in the previous section corresponds to the training probes that are used to fit the logistic regression model.
Composite interval mapping for TIL detected significant peaks on chromosomes 3, 9 and 12 (qGTIL9, qGTIL9 and qGTIL12), explaining 16.1, 9.6 and 20.8 % of the total phenotypic variance, respectively (Fig. 4A and Table 1).
