CHIP-seq流程学习笔记（8）-使用MACS2 bdgdiff提取非生物学重复样本间差异peak进行注释

参考文章：

使用MACS2进行差异peak分析
重点推荐：Call differential binding events

MACS2作为使用最广泛的peak calling软件，在v2版本中添加了差异peak分析的功能，它的子命令功能如下：

1. 软件说明

# 用法说明：
usage: macs2 bdgdiff [-h] --t1 T1BDG --t2 T2BDG --c1 C1BDG --c2 C2BDG
                     [-C CUTOFF] [-l MINLEN] [-g MAXGAP] [--d1 DEPTH1]
                     [--d2 DEPTH2] [--outdir OUTDIR]
                     (--o-prefix OPREFIX | -o OFILE OFILE OFILE)
# 可选参数：
optional arguments:
# 参数 --t1是读取MACS pileup bedGraph for condition 1. 
# 参数 --t2是读取MACS pileup bedGraph for condition 2. 
# 参数 --c1是读取MACS control lambda bedGraph for condition 1.
# 参数 --c2是读取MACS control lambda bedGraph for condition 2.
# 参数 -g 是Maximu gap to merge nearby differential regions.
# 参数 -l Minimum length of differential region. Try bigger value to remove small regions. DEFAULT: 200
# 参数 --d1  Sequencing depth (# of non-redundant reads in million) for condition 1. 
# 参数 --d2  Sequencing depth (# of non-redundant reads in million) for condition 2. 
# 参数 --o-prefix diff_c1_vs_c2保存输出文件名。

2. 操作记录

Step 1: Generate pileup tracks using callpeak module

1. 使用predictd模式对不同样品的片段长度进行预测，以设定相同的extsize大小。
   操作记录如下：

(base) zexing@DNA:~/projects/daizhongye/ChIP_seq/2020_10_29/bam_sort$ macs2 predictd -i Scr.bam.sort
INFO  @ Fri, 06 Nov 2020 16:24:45: # read alignment files...
INFO  @ Fri, 06 Nov 2020 16:24:45: # read treatment tags...
INFO  @ Fri, 06 Nov 2020 16:24:45: Detected format is: BAM
INFO  @ Fri, 06 Nov 2020 16:24:45: * Input file is gzipped.
INFO  @ Fri, 06 Nov 2020 16:24:50:  1000000
INFO  @ Fri, 06 Nov 2020 16:24:56:  2000000
INFO  @ Fri, 06 Nov 2020 16:25:01:  3000000
INFO  @ Fri, 06 Nov 2020 16:25:07:  4000000
INFO  @ Fri, 06 Nov 2020 16:25:12:  5000000
INFO  @ Fri, 06 Nov 2020 16:25:16: 5565504 reads have been read.
INFO  @ Fri, 06 Nov 2020 16:25:16: tag size is determined as 150 bps
INFO  @ Fri, 06 Nov 2020 16:25:16: # tag size = 150
INFO  @ Fri, 06 Nov 2020 16:25:16: # total tags in alignment file: 5565504
INFO  @ Fri, 06 Nov 2020 16:25:16: # Build Peak Model...
INFO  @ Fri, 06 Nov 2020 16:25:16: #2 looking for paired plus/minus strand peaks...
INFO  @ Fri, 06 Nov 2020 16:25:19: #2 number of paired peaks: 75474
INFO  @ Fri, 06 Nov 2020 16:25:19: start model_add_line...
INFO  @ Fri, 06 Nov 2020 16:25:19: start X-correlation...
INFO  @ Fri, 06 Nov 2020 16:25:19: end of X-cor
INFO  @ Fri, 06 Nov 2020 16:25:19: # finished!
INFO  @ Fri, 06 Nov 2020 16:25:19: # predicted fragment length is 275 bps
INFO  @ Fri, 06 Nov 2020 16:25:19: