Too few ovarlapped peaks between MACS2 and Tiles500

[jiangpuxuan]

I ran my data and check the converse of percent of TileMatrix-called peaks that are overlapped by MACS2-called peaks, just like the tutorial 10.3.1.
There is 101547 peaks found by Tiles, 74562 by MACS2, but only 40098 overlaps.
I resize() the Tile to 1000, but Tiles still found the same 101547 peaks.
How could I interpret this situation?

Here is my archR project:

class: ArchRProject 
outputDirectory: /Array-Tuqiang/pxjiang/analysis/atac/CRC_P07A/proj_crc07_hg19 
samples(1): Human_0721
sampleColData names(1): ArrowFiles
cellColData names(34): Sample TSSEnrichment ... ReadsInPeaks FRIP
numberOfCells(1): 1364
medianTSS(1): 6.142
medianFrags(1): 3717

Thank you for your help!

[rcorces]

I would not use the Tiles method for peak calling. I'm making this more clear in the next documentation release. It was really only provided for people who had issues installing MACS2 and it has not been extensively tested.

[jiangpuxuan]

I would not use the Tiles method for peak calling. I'm making this more clear in the next documentation release. It was really only provided for people who had issues installing MACS2 and it has not been extensively tested.

Thank you for your help!