Comments (4)
Is this just a single-end BAM file? I made some modifications to how paired
end files are handled but I didn't think they were invoked unless you
specifically specified -t BAMPE. I believe the default duplicate behavior
is figure out how many duplicates there would be according to the binomial
distribution (so more than one for very high depth of sequencing) and allow
at most that many. You can change that to 1 manually. I'm not clear on why
you are using --nomodel without specifying the shift parameters. Peaks are
filtered by pvalue, requiring min width etc. If you just want to create a
pileup and filter by height, I suggest using bedtools genomeCoverageBed
(the "-d" flag will generate a pileup)
Ben
On Sun, Mar 31, 2013 at 7:22 PM, Ariel Paulson [email protected]:
Perhaps this is related to issue 12, noninteger pileup height...
I have a set of peak calls from the following command (CentOS6 / Python
2.7.2):macs2 callpeak -t IP.bam -c input.bam -g mm -p 0.05 --nomodel
And in the resulting peaks.xls file, the pileup heights are, almost
without exception, higher than the actual pileup heights from the bam file.Using the reported peak chrom, start, and end, I assay bam pileup heights
with the following command:samtools depth -r chrom:start-end IP.bam | cut -f3 | sort | uniq -c
And the maximum is always lower than the macs2 pileup value. This is
especially problematic because I am using pileup heights to filter peaks,
and if I want peaks with height >= 10 then macs2 should not tell me it is
10 when in fact it is 4, or 15 when it is 9, or 22 when 12, etc.Also, I am testing the original bam, and macs2 supposedly removes
duplicates, so the disparities are probably even worse.So how are the pileup values generated, and why are they not the IP height?
Thanks,
Ariel—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18
.
from macs.
Hi Ben,
Just a single-end BAM. I was using --nomodel because of some earlier strangeness with the called peaks (in a different sample) which --nomodel seemed to resolve, so I left it out for the others. But if it is still locally shifting read piles, then this would explain why it is getting pileups which are higher than from the bam directly. And also why the macs:bam pileup ratio seems centered around 2.
Thanks,
Ariel
From: Benjamin Schiller [[email protected]]
Sent: Sunday, March 31, 2013 9:50 PM
To: taoliu/MACS
Cc: Paulson, Ariel
Subject: Re: [MACS] incorrect reported pileup height reported (#18)
Is this just a single-end BAM file? I made some modifications to how paired
end files are handled but I didn't think they were invoked unless you
specifically specified -t BAMPE. I believe the default duplicate behavior
is figure out how many duplicates there would be according to the binomial
distribution (so more than one for very high depth of sequencing) and allow
at most that many. You can change that to 1 manually. I'm not clear on why
you are using --nomodel without specifying the shift parameters. Peaks are
filtered by pvalue, requiring min width etc. If you just want to create a
pileup and filter by height, I suggest using bedtools genomeCoverageBed
(the "-d" flag will generate a pileup)
Ben
Ben Schiller
415.273.9236 (Cell)
UC San Francisco, PhD Candidate
Harvey Mudd College, '08
"There's no such thing as an ordinary human." -The Doctor
On Sun, Mar 31, 2013 at 7:22 PM, Ariel Paulson [email protected]:
Perhaps this is related to issue 12, noninteger pileup height...
I have a set of peak calls from the following command (CentOS6 / Python
2.7.2):macs2 callpeak -t IP.bam -c input.bam -g mm -p 0.05 --nomodel
And in the resulting peaks.xls file, the pileup heights are, almost
without exception, higher than the actual pileup heights from the bam file.Using the reported peak chrom, start, and end, I assay bam pileup heights
with the following command:samtools depth -r chrom:start-end IP.bam | cut -f3 | sort | uniq -c
And the maximum is always lower than the macs2 pileup value. This is
especially problematic because I am using pileup heights to filter peaks,
and if I want peaks with height >= 10 then macs2 should not tell me it is
10 when in fact it is 4, or 15 when it is 9, or 22 when 12, etc.Also, I am testing the original bam, and macs2 supposedly removes
duplicates, so the disparities are probably even worse.So how are the pileup values generated, and why are they not the IP height?
Thanks,
Ariel—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18
.
—
Reply to this email directly or view it on GitHubhttps://github.com//issues/18#issuecomment-15702463.
from macs.
Related to issue #53?
from macs.
Closing due to lack of activity.
from macs.
Related Issues (20)
- Bug: pip install error: subprocess-exited-with-error HOT 2
- What is the difference between "absolute peak summit" and "summit position" in narrowPeak format? HOT 1
- Q: conversion of output files to UCSC track hub format? HOT 1
- Feat: Enhanced clarification of the specific definition of BEDPE as used by MACS HOT 1
- Feat: Reduction of memory consumption HOT 3
- Q: HMMRATAC output and options for _summit.bed HOT 1
- Feat: Built-in scoring for HMMRATAC HOT 1
- Q: HMMRATAC producing too many peaks HOT 10
- Bug: undefined symbol: __pow_finite HOT 3
- Q: HMMRATAC reproducibility HOT 2
- Q: how to use MACS3 for ATAC seq with "himmratac" option? HOT 2
- Bug: Callvar throws an error when run on broadPeak file HOT 15
- Q: impact of sequencing throughput on peak calling HOT 4
- Setting cut-off when using logFE bigWig files from bdgcmp
- Q: How to use HMMRATAC to give us a narrow peak file and have signalValue, p and q-value ? HOT 1
- Bug: OverflowError when running hmmratac
- Number of de-duplicated reads indicated by filterdup log does not match number of lines in output BED file.
- Feat: Add --call-summits option in MACS3 bdgpeakcall?
- Q: using macs3 to detect peaks on long-read readings
- MACS3 callpeak does not create UCSC Browser ready peak files HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from macs.