bigwigs to reads in peaks

Question

[pt.taklifi]

Hello everyone

I have bigwig files of normalized Tn5 insertions . I also have atac seq peaks of the same samples. I was wondering if it is possible to get coverages for atac seq peaks in R using only bigwigs and peaks.

bigwig sample:

"seqnames"  "start" "end"   "width" "strand"    "score"
"1" "chr1"  1   9999    9999    "*" 0
"2" "chr1"  10000   10099   100 "*" 17.7165222167969
"3" "chr1"  10100   10199   100 "*" 30.6012668609619
"4" "chr1"  10200   10299   100 "*" 9.66355800628662
"5" "chr1"  10300   10399   100 "*" 4.83177900314331
"6" "chr1"  10400   10499   100 "*" 8.05296516418457
"7" "chr1"  10500   10699   200 "*" 3.22118592262268
"8" "chr1"  10700   13199   2500    "*" 0

ATAC-seq peaks:

seqnames    start   end name    score   annotation  percentGC   percentAT
chr1    975451  975952  BRCA_39 1.87842575038562    3' UTR  0.6187624750499 0.3812375249501
chr1    1014228 1014729 BRCA_55 4.07469686212787    3' UTR  0.62874251497006    0.37125748502994
chr1    1290080 1290581 BRCA_123    2.44358820293876    3' UTR  0.678642714570858   0.321357285429142
chr1    1291099 1291600 BRCA_124    3.18019908767794    3' UTR  0.702594810379242   0.297405189620758
chr1    1291742 1292243 BRCA_125    8.26783029566134    3' UTR  0.640718562874252   0.359281437125749
chr1    1327977 1328478 BRCA_143    1.08246502080444    3' UTR  0.676646706586826   0.323353293413174
chr1    1334423 1334924 BRCA_151    3.70277788120318    3' UTR  0.634730538922156   0.365269461077844
chr1    1335198 1335699 BRCA_152    2.60759091543721    3' UTR  0.588822355289421   0.411177644710579
chr1    1352725 1353226 BRCA_166    12.7576509548536    3' UTR  0.612774451097804   0.387225548902196

Here is how bigwigs were constructed:

we constructed bigwigs based on the Tn5 offset-corrected insertion sites. To do this, the genome was binned into 100-bp intervals using “tile” in GenomicRanges of the chromosome sizes in R. The insertion sites (GenomicRanges) were then converted into a coverage run-length encoding using “coverage”. Then, to determine the number of Tn5 insertions within each bin we constructed a “Views” object and calculated the sum in each bin with “ViewSums”. We then normalized the total number of reads by a scale factor that converted all samples to a constant 30 million reads within peaks. This approach simultaneously normalizes samples by their quality and read depth, analogous to the reads in peaks normalization within a counts matrix. This was then converted into a bigwig using rtracklayer “export.bw” in R.

[rpolicastro]

Can you explain in more detail what you want to do?

[pt.taklifi]

I added some explanation on how bigwigs were constructed. my objective is to get normalized reads in my atac seq peaks .

[jordi.planells]

Count the reads with featureCounts.

[pt.taklifi]

can you please explain more?

[jordi.planells]

You can treat the peaks as regions and count the reads with whichever tool you like (I said featureCounts because is quite standard). With the reads obtained, you can do the normalization you like; RPKM, TMMs, etc.