一.流程介绍
后续会开发lncRNA、smallRNA、circRNA流程
首先是lncRNA流程,主要分析模块为:
- 测序数据质控;
- 序列比对分析;
- 转录本组装;
- lncRNA分析;
- 表达量分析;
- 表达量差异分析;
- lncRNA靶基因分析;
- 富集分析
1.使用基础ubuntu镜像:
docker run -itd --name lnc ubuntu:18.04
2.安装json环境配置环境
换源:
docker cp sources.list 1a73843e30e5:/etc/apt
apt-get update
apt-get upgrade
lncRNA流程前几个模块与有参流程模块脚本基本一样,软件使用相同,直接复制原来的软件到lnc环境中,并安装即可。
3.测试脚本
01-qc.sh、01-qcStat.sh、02-alignHST.sh、02-alignSummary.sh
使用FastQC质控与Hisat2比对;
使用stringtie进行转录本拼接与组装,分为得到mRNA序列与注释文件与lncRNA序列与注释文件
03-assemblyStie.sh脚本:
for sample in $samplenames
do
delay stringtie
cd $sample
mkdir AssemblyStielncRNA
cd AssemblyStielncRNA
algn_bam=../AlignmentHST/$sample.bam
echo "Transcript assembly for $sample at $(date)"
$stringtie -p $stringtie_threads -G $lnc_gtf \
-o transcripts.gtf -A $sample.exp -l $sample $algn_bam &
cd ../
mkdir AssemblyStieRNA
cd AssemblyStieRNA
algn_bam=../AlignmentHST/$sample.bam
echo "Transcript assembly for $sample at $(date)"
$stringtie -p $stringtie_threads -G $genome_gtf \
-o transcripts.gtf -A $sample.exp -l $sample $algn_bam &
cd ../../
done
根据lncRNA注释文件与mRNA注释文件,拼接得到每个样本的转录本注释文件
mkdir MergedGTFlncRNA_Stie
find */AssemblyStielncRNA/ -name "transcripts.gtf" -type f -printf "%p \n" >transcripts_file-lncRNA.list
$stringtie --merge -p $stringtie_threads -G $lnc_gtf -o MergedGTFlncRNA_Stie/mergedlncRNA.gtf transcripts_file-lncRNA.list
# #$gffcompare -r $genome_gtf -G -o MergedGTFlncRNA_Stie/merged_compare MergedGTFlncRNA_Stie/mergedlncRNA.gtf
mkdir MergedGTFRNA_Stie
find */AssemblyStieRNA/ -name "transcripts.gtf" -type f -printf "%p \n" >transcripts_file-mRNA.list
$stringtie --merge -p $stringtie_threads -G $genome_gtf -o MergedGTFRNA_Stie/mergedmRNA.gtf transcripts_file-mRNA.list
$cufflinks/gffread -w MergedGTFlncRNA_Stie/transcripts-lncRNA.fa -g $genome_fasta MergedGTFlncRNA_Stie/mergedlncRNA.gtf &
$cufflinks/gffread -w MergedGTFRNA_Stie/transcripts-mRNA.fa -g $genome_fasta MergedGTFRNA_Stie/mergedmRNA.gtf &
分别得到所有样本lncRNA与mRNA总注释文件,下一步计算表达量
for sample in $samplenames
do
delay stringtie
cd $sample
mkdir ExpressionStielncRNA
cd ExpressionStielncRNA
algn_bam=../AlignmentHST/$sample.bam
echo "Start estimating for $sample at $(date)"
$stringtie -e -B -p $stringtie_threads -G ../../MergedGTFlncRNA_Stie/mergedlncRNA.gtf \
-o $sample.gtf -A $sample.exp $algn_bam &
cd ../
mkdir ExpressionStieRNA
cd ExpressionStieRNA
algn_bam=../AlignmentHST/$sample.bam
echo "Start estimating for $sample at $(date)"
$stringtie -e -B -p $stringtie_threads -G ../../MergedGTFRNA_Stie/mergedmRNA.gtf \
-o $sample.gtf -A $sample.exp $algn_bam &
cd ../../
done
得到结果如下:
4.接下来使用DEseq2或edger进行差异分析,可参考有参流程分析:
(1)有生物学重复的同时满足 log2FoldChange 绝对值大于等于 1 且 pvalue 小于 0.05 的基因 significant 标为 yes,否则标为 no;(2)没有生物学重复的同时满足 log2FoldChange 绝对值大于等于 1 且 pvalue 小于 1 的基因 significant 标为 yes,否则标为 no。同时我们会根据初始运算结果,对差异阈值做进一步定义或者给予建议,例如:1. fold change 阈值更改;2. 统计检验阈值(pvalue)更改。满足显著性差异阈值的差异基因在分析结果中通过 significant 参数一列为 yes 显现。
5.使用clusterProfiler包 进行KEGG与GO富集分析;
- lncRNA靶基因分析;
10.23更新:
LncRNA预测分析:
使用软件:
CNCI;
CPC2;
LGC;
CNCI
CNCI是中科院计算所赵屹团队开发的一款从转录组中分析编码RNA和非编码RNA的软件。赵屹团队在非编码RNA领域做了很多出色的工作,建立了目前最权威的非编码RNA数据库NONCODE。
github地址:
https://github.com/www-bioinfo-org/CNCI
安装:
git clone [git@github.com](mailto:git@github.com):www-bioinfo-org/CNCI.git
cd CNCI
unzip libsvm-3.0.zip
cd libsvm-3.0
make
cd ..
或者下载到服务器,unzip解压
基本命令为:
python CNCI_package/CNCI.py -f novel.fasta -o CNCI_out -m ve -p 4
参数说明:
-f 输入fasta文件(可以使用-g参数输入GTF文件,但是同时需要使用-d参数指定参考基因组的目录)
-o 输出结果目录
-m 指定模式,脊椎动物选择ve,植物选择pl
-p 指定CPU核数
更多用法参看帮助文档
结果文件:
CPC2
CPC2 下载安装
https://github.com/gao-lab/CPC2_standalone/releases/tag/v1.0.1
安装:
gzip -dc CPC2-beta.tar.gz | tar xf -
cd CPC2-beta
export CPC_HOME="$PWD"
cd libs/libsvm
gzip -dc libsvm-3.18.tar.gz | tar xf -
cd libsvm-3.18
make clean && make
cd $CPC_HOME
bin/CPC2.py -i (input_seq) -o (result_in_table)
运行命令:
python3 CPC2.py -i transcripts-lncRNA.fa
结果展示:
LGC
LGC是由北京基因组所基于python2 开发的一款快速lncRNA预测工具,该工具通过ORF(开放阅读框)长度和GC含量间的关系进行相关运算来鉴定lncRNA。LGC最大特点是能够基于跨物种策略进行lncRNA发掘。因此LGC可以支持有参数据和无参数据 进行lncRNA鉴定。在区分从植物到哺乳动物的不同物种的lncRNA和蛋白编码RNA方面,LGC鉴定的准确率高达90%。
下载与安装:
wget [http://bigd.big.ac.cn/biocode/tools/4/releases/1.0/file/download?filename=lgc-1.0.tar.gz](http://bigd.big.ac.cn/biocode/tools/4/releases/1.0/file/download?filename=lgc-1.0.tar.gz)
tar zxf lgc-1.0.tar.gz
chmod 755 lgc-1.0.py
#确保conda,lgc-1.0.py在环境变量里
lgc-1.0.py input.fasta output.txt
# Or
python lgc-1.0.py input.fasta output.txt
结果展示:
二.nextflow使用
这里使用 Nextflow 作为流程搭建工具,它有着很多强大的功能,第一次接触,首先先安装:
wget -qO- [https://get.nextflow.io](https://get.nextflow.io/) | bash
或者使用conda安装
conda install nextflow
官方参考文档:
https://www.nextflow.io/docs/latest/getstarted.html
这里可以直接使用
直接使用已有lncRNA流程进行测试,参考地址
或者使用下面代码使用rna流程:
nextflow pull nf-core/rnaseq
下载完流程之后:
运行文件为nf结尾文件,
文件内容如下:
#!/usr/bin/env nextflow
import SampleGroup
import Config
/*------------------RNA分析nextflow-pipeline初版-----------------------------
RNA_np_v0.0.1
System: Linux version 3.10.0-693.21.1.el7.x86_64
Tool : Nextflow 0.30.2.4867
-------------------------------------------------------------------------*/
/*---------------------------定义基本目录-----------------------------------
base pathway
database: 与人hg19相关的数据、工具存储的目录
opd: output pathway本项目的输出文件母目录
-------------------------------------------------------------------------*/
database="/data/zhangxc/database/homo_sapiens"
opd = "/data/zhangxc/lnc_rna/lncRNA/output"
main_dir = "/data/zhangxc/lnc_rna/lncRNA"
/*-------------------------读取样本及定义组---------------------------------
sample and group
data.ini为样本和组的配置文件,通过congfig.groovy将定义好的样本和组读入pipline中备用
sample = {{sample_name},{sample_file1,sample_file2}}
group = {{group_name},{control_name},{case_name},{control_name,case_name}}
nextflow 同一命名变量只允许一次作为输入变量
sample 扩展成3份 分别用来进行fastqc、SOAPnuke、println
-------------------------------------------------------------------------*/
params.data_file="/data/zhangxc/lnc_rna/lncRNA/test/data.ini"
config=new Config(params.data_file)
Channel.from( config.ReadSamples() ).set{samples}
samples.into{samples0;samples1;sample_2}
Channel.from( config.ReadGroups() ).set{groups0}
sample_2.println()
/*------------------------------fastqc------------------------------------
fastqc
质量控制模块,独立于分析流程的单独模块
input : sample
output: otp/fastqc/sample_name
-------------------------------------------------------------------------*/
process fastqc{
tag { sample_name }
// container 'biocontainers/fastqc'
// conda "fastqc"
publishDir { "output/fastqc/"+ sample_name }
input:
set sample_name , files from samples0
output:
file "*_fastqc/Images/*.png"
script:
if(files.size == 1){
"""
echo \$PWD
fastqc --extract -o . ${files[0]}
"""
}else{
"""
echo \$PWD
fastqc --extract -o . ${files[0]} ${files[1]}
"""
}
}
/*------------------------------soapnuke------------------------------------
soapnuke
质量控制模块,开始与分析第一步,用来去除引物生成clean sample
input : sample
output: file otp/soapnuke/sample_name/*.fq.gz val clean_samples-->{{sample_name},{*.fq.gz}}
clean_samples copy to clean_samples1|clean_samples2
-------------------------------------------------------------------------*/
process soapnuke{
tag { sample_name }
// container 'registry.cn-hangzhou.aliyuncs.com/bio/soapnuke'
publishDir path:{"output/soapnuke/"+sample_name} ,mode:'copy',
saveAs: {filename ->
if (filename.indexOf(".txt")>0) filename
else null
}
input:
set sample_name , files from samples1
output:
set sample_name , file("*.fq.gz") into clean_samples
file "*.txt"
script:
"""
echo \$PWD
SOAPnuke filter -1 ${files[0]} -2 ${files[1]} -l 20 -q 0.5 \
-f GATCGGAAGAGCACACGTCTGAACTCCAGTCAC -r GATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT \
-o ./ -C ${sample_name}_1.clean.fq.gz -D ${sample_name}_2.clean.fq.gz
"""
}
clean_samples.into{clean_samples1;clean_samples2;}
/*------------------------------hisat------------------------------------
hisat
比对软件,与bowtie和tophat功能相似,将read与基因组作比对
input : clean_samples1
output: file otp/hisat/sample_name/${sample_name}.sam val sam-->{{sample_name},{${sample_name}.sam}}
file otp/hisat/sample_name/*.align_summary.txt val summary_txt-->{{sample_name},{*.align_summary.txt}}
-------------------------------------------------------------------------*/
process hisat{
tag { sample_name }
// container 'registry.cn-hangzhou.aliyuncs.com/bio/hisat2'
publishDir { "output/hisat/"+ sample_name }
input:
set sample_name , files from clean_samples1
output:
set sample_name , file("${sample_name}.sam") into sam
set sample_name , file("*.align_summary.txt") into summary_txt
script:
"""
echo \$PWD
hisat2 -x $database/grch37_tran/genome_tran -1 ${files[0]} -2 ${files[1]} -S ${sample_name}.sam 2> ${sample_name}.align_summary.txt
"""
}
/*------------------------------samtools----------------------------------
samtools
工具软件,将sam转为bam格式,并根据染色体定位进行sort和index
input : val sam
output: file otp/samtools/sample_name/${sample_name}.bam val sam-->{{sample_name},{${sample_name}.bam}}
-------------------------------------------------------------------------*/
process samtools{
tag { sample_name }
// container 'broadinstitute/genomes-in-the-cloud:2.3.1-1512499786'
publishDir { "output/samtools/"+ sample_name }
input:
set sample_name , file('sample.sam') from sam
output:
set sample_name , file("${sample_name}.bam") into bam
script:
"""
echo \$PWD
samtools view -bS sample.sam > ${sample_name}_unsorted.bam
samtools sort -@ 8 ${sample_name}_unsorted.bam ${sample_name}
samtools index ${sample_name}.bam ${sample_name}.bam.bai
"""
}
/*------------------------------htseq----------------------------------
htseq
转录组表达量分析
input : val bam
output: file otp/htseq/sample_name/${sample_name}.rawCount.txt val htseq_txt-->{{sample_name},{${sample_name}.rawCount.txt}}
htseq_txt copy to htseq_txt0|htseq_txt1|htseq_txt2
-------------------------------------------------------------------------*/
process htseq{
tag { sample_name }
//container 'broadinstitute/genomes-in-the-cloud:2.3.1-1512499786'
//container "dyndna/docker-for-pcawg14-htseq"
// conda "htseq=0.9.1"
publishDir { "output/htseq/"+ sample_name }
input:
set sample_name , file('bam_files') from bam
output:
set sample_name , file("${sample_name}.rawCount.txt") into htseq_txt
script:
"""
echo \$PWD
htseq-count -f bam -m union -s yes -t exon -i gene_id -r pos bam_files \
$database/Homo_sapiens.GRCh37.75.gtf >${sample_name}.rawCount.txt
"""
}
htseq_txt.into{htseq_txt0;htseq_txt1;htseq_txt2}
htseq_txt2.println()
/*------------------------------htseq_xls----------------------------------
htseq_xls
转录组表达量分析后,将txt文件转换格式为xls
input : val summary_txt
val htseq_txt0
output: file otp/htseq_xls/sample_name/${sample_name}.rpkm.xls val htseq_xls-->{{sample_name},{${sample_name}.rpkm.xls}}
-------------------------------------------------------------------------*/
process htseq_xls{
tag { sample_name }
//container 'registry.cn-hangzhou.aliyuncs.com/bio/hisat2'
// container "broadinstitute/genomes-in-the-cloud:2.3.1-1512499786 "
publishDir { "output/htseq/"+ sample_name }
input:
set sample_name , file('txt_files') from summary_txt
set sample_name , file('sample.rawCount.txt') from htseq_txt0
output:
set sample_name , file("${sample_name}.rpkm.xls") into htseq_xls
script:
"""
echo \$PWD
perl /data/zhangxc/lnc_rna/lncRNA/bin/RNAseq_bin/htseq2rpkm_hisat.pl txt_files \
$database/hg19.GRCh37.74.ncRNA.gene.len \
sample.rawCount.txt $sample_name ${sample_name}.rpkm.xls
sed 's/^/$sample_name\t/' ${sample_name}.rpkm.xls | sed '1d' > ${sample_name}.rpkm.tmp
"""
}
只截取到基因表达量部分,感兴趣的可以去尝试运行,运行方式为:
nextflow run zxc.nf