Galaxy thinks gd7 fastq all have low quality scores (1 to 4 on a scale of -5 to 40).
Mapping .gff gd7 reads to browser to check quality. Looks good on UCSC browser, except that squish display doesn’t work.
Taking +/- 100 bp of promoter to ID paused promoters. Map to genome (dm2), use count to convert to scores per gene. Align to tiling data.
Take c1>1000 (raw reads aligned to promoter) and c3-c4 > 5 (raw tiling array 5 greater in toll10b than gd7)