|Alternative names||Model-based Analysis for ChIP-Seq, MACS2, macs2|
|Interface||Command line interface|
|Restrictions to use||None|
|Input data||A tag file, a treatment file.|
|Output data||A file with information about called peaks, peak locations, peak summits locations, negative peaks, a script to produce a PDF image, a file that can be viewed through the UCSC genome browser, a diagnosis report and an optional file for the subpeaks option.|
|Programming languages||C, Python|
|License||BSD 3-clause “New” or “Revised” License|
|Requirements||Numpy, GCC, Cython|
Add your version
- Issues: https://github.com/taoliu/MACS/
- person_outline Wei Li <>
- person_outline Tao Liu <>
#1 opened on 2015-12-24 by Arnaud Desfeux • 1 answer
new:~/sigma $ macs14 -t Alice1_S7.sam --format=SAM INFO @ Wed, 11 Feb 2015 14:34:25: # ARGUMENTS LIST: # name = NA # format = SAM # ChIP-seq file = Alice1_S7.sam # control file = None # effective genome size = 2.70e+09 # band width = 300 # model fold = 10,30 # pvalue cutoff = 1.00e-05 # Large dataset will be scaled towards smaller dataset. # Range for calculating regional lambda is: 10000 bps INFO @ Wed, 11 Feb 2015 14:34:25: #1 read tag files... INFO @ Wed, 11 Feb 2015 14:34:25: #1 read treatment tags... Traceback (most recent call last): File "/usr/bin/macs14", line 366, in <module> main() File "/usr/bin/macs14", line 60, in main (treat, control) = load_tag_files_options (options) File "/usr/bin/macs14", line 335, in load_tag_files_options ttsize = tp.tsize() File "/usr/lib/python2.6/site-packages/MACS14/IO/Parser.py", line 655, in tsize return int(s/n) ZeroDivisionError: integer division or modulo by zero
Publication for Model-based Analysis for ChIP-Seq
[…] hg38 reference genome using bowtie2 (version 2.3.1) (langmead and salzberg, 2012) using default parameters except that up to one mismatch was accepted in the seed sequence. peaks were identified by macs2 (version 2.1.1) (zhang et al., 2008) using the input sample as a control while bypassing the shifting model and with the broad peak setting with a broad cutoff of 0.05 for h3k4me3 and h3k79me2 […]
[…] (hg19) human reference sequence using bwa-mem (li and durbin, 2009). after alignment, technical replicates were merged and all further analyses were performed on these merged data. for peak calling, macs2 (zhang et al., 2008) was used with no-model, 100-bp shift, 200-bp extension, and broad peaks options. only peaks called with a peak score (q-value) of 1% or better were kept from each sample, […]
[…] default parameters. the reads were mapped to the reference genome using bowtie v2 2.1.0. the non-unique reads were randomly distributed. binding enrichment was called from the aligned reads using macs2 2.0.10 and spp (http://compbio.med.harvard.edu/supplements/chip-seq) using default parameters. all statistical analyses were performed using r., the 5΄ end hypermethylated spliced leader (sl) […]
[…] respectively, for peak-calling from the same sequencing depth in all samples65. significant enrichment regions in h3k4me3 and h3k27me3 samples, relative to h3 control samples, were identified using macs2 2.1.0 (default settings for h3k4me3—fdr<0.05; broad-peaks mode for h3k27me3—fdr<0.1)6667. regions with signal level measured by fpkm(chip)/fpkm(h3 control) differing between biological […]
[…] reads were subsequently filtered for alignment quality of > q30 and were required to be properly paired. reads mapping to the mitochondria or chromosome y were removed and not considered. we used macs2 (http://pypi.python.org/pypi/macs2) to call all reported atac-seq peaks. macs2 was used with the following parameters (--nomodel --shift 0). peaks were filtered using the consensus excludable […]
I think it's important to understand that it is designed for point-source data, and for that it works perfectly (that's why I gave maximum rating, despite the constraints described in the following section, which I don't see as a fault - it's really not intended for the tasks below.) Numerous tests show (even a paper published by the MACS authors) that when it comes to diffuse enrichments, most notably, ChIP-seq with histone marks, then MACS fails to identify the peaks correctly. Usually it manages to detect some parts of it, some local summits within the dispersed signal, if it detects something at all. Although in the literature it is used for various datasets, including histone marks, ATAC-seq, MeDIP-seq etc., I wouldn't recommend it for detecting other enrichments than point-source peaks. There are much superior tools out there for diffuse signals (like SICER). I have to mention though that maybe H3K4me3 is an exception, it is a very "TF-like" histone mark, with sharp, clear, high enrichments, and for that mark only MACS might be acceptable. But I wouldn't trust it for any other HMs (and I've been working with ChIP-seq data of all histone marks you can imagine for ~10 years). I would also take the results of MeDIP-seq. ATAC-seq and similar data with a grain of salt if it was processed with MACS, because while it can certainly detect some enrichment, not all ofthem can be considered point-source, many methylated regions, open chromatin etc. are way broader than that.
With that said, I think it's a great tool for what it was designed.
Be aware that the MACS website is still saying that the newest version is version 1.4.2, which is not true, it has long been superceded by the major version 2, which you can find on GitHub (at the time of writing this review the latest stable version is 2.1.0).
Easy to install and use.