对于strand-specific的RNA-seq而言,我们必须得知道它是哪一种建库方式,才能进行后续的定量分析。
stringtie:
--rf Assumes a stranded library fr-firststrand.
--fr Assumes a stranded library fr-secondstrand.
kallisto:
--fr-stranded runs kallisto in strand specific mode, only fragments where the first read in the pair pseudoaligns to the forward strand of a transcript are processed. If a fragment pseudoaligns to multiple transcripts, only the transcripts that are consistent with the first read are kept.
--rf-stranded same as --fr-stranded but the first read maps to the reverse strand of a transcript.
现在比较常用的方式是fr-firststrand,也就是基于d-UTP的建库方式。但是为了更稳妥的判断,我们可以使用RSeQC中的工具来判断。RSeQC是2012年发表在Bioinformatics上的一个工具,包含多种功能:
1. 安装
# pip安装
pip3 install RSeQC
# 源代码安装
tar zxf RSeQC-VERSION.tar.gz
cd RSeQC-VERSION
#type 'python setup.py install --help' to see options
python setup.py install #Note this requires root privilege
or
python setup.py install --root=/home/user/XXX/ #install RSeQC to user specificed location, does NOT require root privilege
#This is only an example. Change path according to your system configuration
export PYTHONPATH=/home/user/lib/python2.7/site-packages:$PYTHONPATH
#This is only an example. Change path according to your system configuration
export PATH=/home/user/bin:$PATH
2. infer_experiment.py
单端数据:
infer_experiment.py -r hg19.refseq.bed12 -i SingleEnd_StrandSpecific_36mer_Human_hg19.bam
#Output:
This is SingleEnd Data
Fraction of reads failed to determine: 0.0170
Fraction of reads explained by "++,--": 0.9669
Fraction of reads explained by "+-,-+": 0.0161
"++,--" 的比例远远超过另一种,这是strand-specifc的数据。++,--就是指的测出来的正链即实际的正链,负链就是实际的负链。
如上图这种,就是非链特异性的单端数据。
如果两种接近1:1,则是非链特异性,而假如两种比例悬殊,则是链特异性。
双端数据:
infer_experiment.py -r hg19.refseq.bed12 -i Pairend_StrandSpecific_51mer_Human_hg19.bam
#Output::
This is PairEnd Data
Fraction of reads failed to determine: 0.0072
Fraction of reads explained by "1++,1--,2+-,2-+": 0.9441
Fraction of reads explained by "1+-,1-+,2++,2--": 0.0487
这种显然是链特异性,而且是fr-secondstrand。意思就是read1在+链,相对的gene也同样在+链上,而read2在+链,相对的gene在-链上。这种就是kallisto中的--fr-stranded和stringtie中的--fr。
现在这种特异性的library相对较少,而下面这种更为常见:
主要是“1+-,1-+,2++,2--”这种,也就是read1在+链,相对的gene其实是在-链(reverse)。这种就是“fr-firststrand”,也就是参数中的--rf。
同样两种在0.5附近的是non-specific:
infer_experiment.py -r hg19.refseq.bed12 -i Pairend_nonStrandSpecific_36mer_Human_hg19.bam
#Output::
This is PairEnd Data
Fraction of reads failed to determine: 0.0172
Fraction of reads explained by "1++,1--,2+-,2-+": 0.4903
Fraction of reads explained by "1+-,1-+,2++,2--": 0.4925
判断所需要的refseq文件可以在说明页面找到下载:
参考:
欢迎关注~
网友评论