转录组分析流程

原創 sunchengquan 2019-01-01 20:32

文章目录

分析流程概述

参考文献：hppRNA—a Snakemake-based handy parameter-free pipeline for RNA-Seq analysis of numerous samples

六种主要流程示意图：
在这里插入图片描述

下载测试数据

wget ftp://ftp.ccb.jhu.edu/pub/RNAseq_protocol/chrX_data.tar.gz

如果文章中的数据不是fastq格式，而是给GSExxxx，这是需要下载SRA数据

需要安装sratoolkit，SRR_Acc_List.txt文件储存SRR号码

cat SRR_Acc_List.txt | while read id; do (prefetch  ${id} &);done


# 批量转换sra到fq格式
ls /public/project/RNA/airway/sra/*  | while read id; do ( nohup fastq-dump --gzip --split-3 -O ./ ${id} & ); done

下载的数据：

[sunchengquan 15:45:09 /data/Data_base/test_tmp/RNA_seq_practice/chrX_data/samples]
$ ll
总用量 1.8G
-rwxr-xr-x 1 sunchengquan sunchengquan  88M 1月  15 2016 ERR188044_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  88M 1月  15 2016 ERR188044_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  84M 1月  15 2016 ERR188104_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  85M 1月  15 2016 ERR188104_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan 108M 1月  15 2016 ERR188234_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan 109M 1月  15 2016 ERR188234_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  59M 1月  15 2016 ERR188245_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  60M 1月  15 2016 ERR188245_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  65M 1月  15 2016 ERR188257_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  66M 1月  15 2016 ERR188257_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  39M 1月  15 2016 ERR188273_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  39M 1月  15 2016 ERR188273_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  88M 1月  15 2016 ERR188337_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  88M 1月  15 2016 ERR188337_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  63M 1月  15 2016 ERR188383_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  63M 1月  15 2016 ERR188383_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  86M 1月  15 2016 ERR188401_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  87M 1月  15 2016 ERR188401_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  56M 1月  15 2016 ERR188428_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  56M 1月  15 2016 ERR188428_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  70M 1月  15 2016 ERR188454_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  70M 1月  15 2016 ERR188454_chrX_2.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  75M 1月  15 2016 ERR204916_chrX_1.fastq.gz
-rwxr-xr-x 1 sunchengquan sunchengquan  75M 1月  15 2016 ERR204916_chrX_2.fastq.gz

数据质量控制

reads质量评估软件：fastqc生成质控报告，multiqc将各个样本的质控报告整合为一个。

reads质量控制软件：prinseq，cutadapt，trimmomatic,trim_galore

查看qc.sh


#!/usr/bin/env bash

set -e

settings(){
        samples=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/samples
        if test -w $samples;then
                mkdir -p {$samples/qc,$samples/cleandata/qc}   
        else
                echo "没有写入权限"
        fi

}


thread(){
        tmp_fifofile="/tmp/$$.fifo" #脚本运行的当前进程ID号作为文件名 
    mkfifo "$tmp_fifofile" 
    exec 6<>"$tmp_fifofile"  #将fd6指向fifo类型
    rm $tmp_fifofile
    thread_num=$1 # 此处定义线程数
    for((i=0;i<$thread_num;i++));do 
        echo 
    done >&6 # 事实上就是在fd6中放置了$thread个回车符
    $2 6 $3
    exec 6>&- # 关闭df6
}


qc(){
        source activate RNA
        printf "[%s %s %s %s %s %s]::数据质量评估\n" $(echo `date`)
        start=$(date +%s.%N)
        list=$(find $2 -name *q\.gz)
        file_num=`ls -l $2/qc|wc -l`
        if [ $file_num -lt 2 ];then
                for i in $list;do
                        read -u$1
                        {
                        name=`awk -v each=$i 'BEGIN{split(each,arr,"/");l=length(arr);print arr[l]}' `
                        fastqc  $i -o $2/qc  &>> $2/qc/qc.log
                        printf "[%s %s %s %s %s %s]::%s质量评估完成\n" $(echo `date`) $name
                        echo >&$1 
                        } &
                done && wait

                multiqc -d $2/qc -o $2/qc 
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l) 
                printf "[%s %s %s %s %s %s]::数据质量评估耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}

trim_qc(){
        printf "[%s %s %s %s %s %s]::数据质量控制\n" $(echo `date`)
        source activate RNA
        start=$(date +%s.%N)
        dir=$samples/cleandata
        find $samples -name *1?f*q?gz|sort >$dir/1
        find $samples -name *2?f*q?gz|sort >$dir/2
        paste -d ":" $dir/1 $dir/2  > $dir/config && rm $dir/1 $dir/2
        file_num=`ls -l $dir|wc -l`
        
        if [ $file_num -lt 3 ];then
                for id in `cat $dir/config`;do
                        read -u$1
                        fq1=$(echo $id|cut -d":" -f1)
                        fq2=$(echo $id |cut -d":" -f2)
                        base_name=$(basename $fq1)
                        name=`awk -v each=$base_name 'BEGIN{split(each,arr,"_");print arr[1]}' ` 
                        {
                        trim_galore -q 25 --phred33 --length 25 --stringency 3 --paired -o $dir $fq1 $fq2 &> $dir/trim.log
                        printf "[%s %s %s %s %s %s]::%s质量控制完成\n" $(echo `date`) $name
                        echo >&$1
                        } &
                done && wait 
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::数据质量控制耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}


settings
thread 6 qc $samples
thread 3 trim_qc 
thread 6 qc $samples/cleandata

Tophat –> Cufflink –> Cuffdiff

参考文献：Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks

流程图：
在这里插入图片描述

手动安装相关软件

我们已经使用bioconda安装相关的软件，现在手动安装一下，本流程所需要的软件

下载并安装比对软件bowtie2
cd ~/local/app
curl -OL http://downloads.sourceforge.net/project/bowtie-bio/bowtie2/2.2.4/bowtie2-2.2.4-linux-x86_64.zip
unzip bowtie2-2.2.4-linux-x86_64.zip

把比对软件以及相关程序链接到bin文件夹
ln -s ~/local/app/bowtie2-2.2.4/bowtie2 ~/bin/
ln -s ~/local/app/bowtie2-2.2.4/bowtie2-align* ~/bin/
ln -s ~/local/app/bowtie2-2.2.4/bowtie2-build ~/bin/

安装tophat2
cd ~/local/app/
curl -OL http://ccb.jhu.edu/software/tophat/downloads/tophat-2.1.1.Linux_x86_64.tar.gz
tar zxvf tophat-2.0.13.Linux_x86_64.tar.gz
cd ~/bin/
vi tophat
#!/usr/bin/env bash
python2 ~/local/app/tophat-2.1.1.Linux_x86_64/tophat $@
chmod 755 tophat
保存退出

#注释文件
cd /data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genes
curl -L ftp://ftp.ensembl.org/pub/release-77/gtf/homo_sapiens/Homo_sapiens.GRCh38.77.gtf.gz > hg38.gtf.gz
gunzip *.gz
cat hg38.gtf | awk ' $1 =="X" { print $0 }' > chr_X.gtf


安装cufflinks
cd ~/local/app
curl -OL http://cole-trapnell-lab.github.io/cufflinks/assets/downloads/cufflinks-2.1.1.Linux_x86_64.tar.gz
tar zxvf cufflinks-2.1.1.Linux_x86_64.tar.gz

ln -fs ~/local/app/cufflinks-2.1.1.Linux_x86_64/cufflinks ~/bin
ln -fs ~/local/app/cufflinks-2.1.1.Linux_x86_64/cuffdiff ~/bin
ln -fs ~/local/app/cufflinks-2.1.1.Linux_x86_64/gtf_to_sam ~/bin
ln -fs ~/local/app/cufflinks-2.1.1.Linux_x86_64/cuffcompare ~/bin
cd ~/bin
vi cuffmerge
#!/usr/bin/env bash
python2 ~/local/app/cufflinks-2.1.1.Linux_x86_64/cuffmerge $@
chmod 755 cuffmerge

流程代码

#!/usr/bin/env bash

set -ue

settings(){
        samples=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/samples
        index=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genome/index
        output=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/tophat+cuff
        if test -w $(dirname $output) &&  test -w $(dirname index);then
                mkdir -p {$index/bowtie,$output/1_tophat,$output/2_cufflinks,$output/3_cuffdiff}
        fi
        cuffdiff=$output/3_cuffdiff
        indexes=$index/bowtie/chrX
        genome=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genome/chrX.fa
        gene=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genes/chrX.gtf
}


thread(){
    tmp_fifofile="/tmp/$$.fifo" 
    mkfifo "$tmp_fifofile" 
    exec 6<>"$tmp_fifofile"  
    rm $tmp_fifofile
    thread_num=$1 
    for((i=0;i<$thread_num;i++));do 
        echo 
    done >&6 
    $2 6 
    exec 6>&- 
}


index(){
        printf "[%s %s %s %s %s %s]::建立索引bowtie2-build\n" $(echo `date`)
        start=$(date +%s.%N)
        file_num=`ls -l $index/bowtie|wc -l`    
        source activate RNA
        base_name=$(basename $genome)
    name=`awk -v each=$base_name 'BEGIN{split(each,arr,".");print arr[1]}' ` 
        if [ $file_num -lt 2 ];then
                bowtie2-build -f $genome $index/bowtie/$name &> $index/bowtie/index.log
                ln -s $genome $index/bowtie/$basename
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::建立索引耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}


mapping(){
        printf "[%s %s %s %s %s %s]::与参考基因组比对tophat\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/1_tophat
        find $samples/cleandata -name *1?f*q.gz|sort > $dir/1
        find $samples/cleandata -name *2?f*q.gz|sort > $dir/2
        paste -d ":" $dir/1 $dir/2  > $dir/config && rm $dir/1 $dir/2
        file_num=`ls -l $dir|wc -l`
        source activate RNA
        if [ $file_num -lt 3 ];then
                for id in $(cat $dir/config);do
                        fq1=$(echo $id|cut -d":" -f1)
                        fq2=$(echo $id |cut -d":" -f2)
                        name=`awk -v each=$(basename $fq1) 'BEGIN{split(each,arr,"_");print arr[1]}' ` 
                        read -u$1
                        {
                        tophat -p 8 -G $gene -o $dir/$name $indexes $fq1 $fq2 &>> $dir/mapping.log
                        printf "[%s %s %s %s %s %s]::%s比对完成\n" $(echo `date`) $name
                        echo >&$1
                        } &
                done && wait
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::比对耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}


assemble(){
        printf "[%s %s %s %s %s %s]::转录本组装和定量cufflinks\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/2_cufflinks
        file_num=`ls -l $dir|wc -l`
        source activate RNA
        if [ $file_num -lt 3 ];then
                for id in $(cat $output/1_tophat/config);do
                        fq1=$(echo $id|cut -d":" -f1)
                        name=`awk -v each=$(basename $fq1) 'BEGIN{split(each,arr,"_");print arr[1]}' ` 
                        read -u$1
                        {
                        cufflinks -p 8 -g $gene -o $dir/$name $output/1_tophat/$name/accepted_hits.bam &> $dir/$name.log
                        printf "[%s %s %s %s %s %s]::%s转录本组装完成\n" $(echo `date`) $name
                        echo >&$1
                        } &
                done && wait
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::转录本组装耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA

}


merge(){
        printf "[%s %s %s %s %s %s]::转录本合并cuffmerge\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/2_cufflinks
        find $dir -name *transcripts?gtf|sort > $dir/assemblies.txt
        source activate RNA
        if [ ! -d $dir/merged_asm ];then
                cuffmerge -p 8 -o $dir/merged_asm -g $gene -s $genome $dir/assemblies.txt &> $dir/cuffmerge.log
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::转录本合并耗时%.2f分钟\n" $(echo `date`) $dur
        fi
    source deactivate RNA

}


diff(){
        printf "[%s %s %s %s %s %s]::差异分析cuffdiff\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/3_cuffdiff
        S1=$output/1_tophat/ERR188245/accepted_hits.bam;S2=$output/1_tophat/ERR188428/accepted_hits.bam;S3=$output/1_tophat/ERR188337/accepted_hits.bam
        S4=$output/1_tophat/ERR204916/accepted_hits.bam;S5=$output/1_tophat/ERR188234/accepted_hits.bam;S6=$output/1_tophat/ERR188273/accepted_hits.bam
        S7=$output/1_tophat/ERR188401/accepted_hits.bam;S8=$output/1_tophat/ERR188257/accepted_hits.bam;S9=$output/1_tophat/ERR188383/accepted_hits.bam
        S10=$output/1_tophat/ERR188454/accepted_hits.bam;S11=$output/1_tophat/ERR188104/accepted_hits.bam;S12=$output/1_tophat/ERR188044/accepted_hits.bam
        
        file_num=`ls -l $dir|wc -l`
        source activate RNA
        if [ $file_num -lt 3 ];then
                cuffdiff -p 8 -b $genome -o $dir -L Female,Male -u $output/2_cufflinks/merged_asm/merged.gtf $S1,$S2,$S3,$S4,$S5,$S6 $S7,$S8,$S9,$S10,$S11,$S12 &> $dir/cuffdiff.log
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::差异分析cuffdiff耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}

expression_matrix(){
        dir=$output/3_cuffdiff
        expr=$dir/gene_exp.diff
        ##筛选出下调的基因（log2_fold_change < -2 & pvalue < 0.001）
        awk '{if(($10<-2)&&($11<0.001))print $3"\t"$8"\t"$9"\t"$10}' $dir/gene_exp.diff | grep -v 'inf' > $dir/down.txt
        ## 筛选出上调的基因（log2_fold_change > 2 & pvalue < 0.001
        awk '{if(($10>2)&&($11<0.001))print $3"\t"$8"\t"$9"\t"$10}' $dir/gene_exp.diff | grep -v 'inf' > $dir/up.txt
}


settings
index
thread 4 mapping
thread 4 assemble
merge
diff

HISAT2 ->StringTie -> Ballgown

参考文献:Transcript-level expression analysis of RNA-seq experiments with HISAT, StringTie and Ballgown
流程示意图
在这里插入图片描述

流程代码

#!/usr/bin/env bash
set -ue

settings(){
        samples=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/samples
        index=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genome/index
        output=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/hisat+stringtie
        if test -w $(dirname $output) &&  test -w $(dirname index);then
                mkdir -p {$index/hisat2,$output/1_hisat,$output/2_stringtie,$output/3_ballgown,$output/4_DE}
        fi
        hisat=$output/1_hisat
        stringtie=$output/2_stringtie
        ballgown=$output/3_ballgown
        genome=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genome/chrX.fa
        gene=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genes/chrX.gtf
}

thread(){
    tmp_fifofile="/tmp/$$.fifo" 
    mkfifo "$tmp_fifofile" 
    exec 6<>"$tmp_fifofile"  
    rm $tmp_fifofile
    thread_num=$1 
    for((i=0;i<$thread_num;i++));do 
        echo 
    done >&6 
    $2 6 
    exec 6>&- 
}

index(){
        printf "[%s %s %s %s %s %s]::建立索引hisat2-build\n" $(echo `date`)
        start=$(date +%s.%N)
        file_num=`ls -l $index/hisat2|wc -l`    
        source activate RNA
        base_name=$(basename $genome)
                name=`awk -v each=$base_name 'BEGIN{split(each,arr,".");print arr[1]}' ` 
        if [ $file_num -lt 2 ];then
                                hisat2_extract_exons.py $gene >$index/hisat2/$name'.exon'
                                hisat2_extract_splice_sites.py $gene >$index/hisat2/$name'.ss'
                                hisat2-build --ss $index/hisat2/$name'.ss' --exon $index/hisat2/$name'.exon' $genome $index/hisat2/$name 1>$index/hisat2/index.log 2>&1
                                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::建立索引耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
        
}

mapping(){
        printf "[%s %s %s %s %s %s]::与参考基因组比对hisat2\n" $(echo `date`)
    start=$(date +%s.%N)
    dir=$output/1_hisat
    find $samples/cleandata -name *1?f*q.gz|sort > $dir/1
    find $samples/cleandata -name *2?f*q.gz|sort > $dir/2
    paste -d ":" $dir/1 $dir/2  > $dir/config && rm $dir/1 $dir/2
    file_num=`ls -l $dir|wc -l`
        index_prefix=`awk -v each=$(basename $genome) 'BEGIN{split(each,arr,".");print arr[1]}' `
    source activate RNA
    if [ $file_num -lt 3 ];then
        for id in $(cat $dir/config);do
            fq1=$(echo $id|cut -d":" -f1)
            fq2=$(echo $id |cut -d":" -f2)
            name=`awk -v each=$(basename $fq1) 'BEGIN{split(each,arr,"_");print arr[1]}' ` 
            read -u$1
            {
                        hisat2 -p 8 --dta -x $index/hisat2/${index_prefix} -1 $fq1 -2 $fq2 -S $dir/${name}.sam &> $dir/${name}.hisat.log
                        samtools sort -@ 8 -o $dir/${name}.bam $dir/${name}.sam &> /dev/null && rm $dir/${name}.sam 
            printf "[%s %s %s %s %s %s]::%s比对完成\n" $(echo `date`) $name
            echo >&$1
            } &
        done && wait
        dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::比对耗时%.2f分钟\n" $(echo `date`) $dur
    fi
    source deactivate RNA

}

assemble(){
        printf "[%s %s %s %s %s %s]::转录本组装stringtie\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/2_stringtie
        file_num=`ls -l $dir|wc -l`
        source activate RNA
        if [ $file_num -lt 3 ];then
                for id in $output/1_hisat/*.bam;do
                        base_name=$(basename $id)
                        i=${base_name%.bam*}
                        read -u$1
                        {
                        stringtie -p 8 -G $gene -o $dir/${i}.gtf -l $i $output/1_hisat/${i}.bam &>> $dir/assemble.log
                        printf "[%s %s %s %s %s %s]::%s转录本组装完成\n" $(echo `date`) $i
                        echo >&$1
                        } &
                done && wait
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::转录本组装耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}


merge(){
        printf "[%s %s %s %s %s %s]::转录本合并stringtie --merge\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/2_stringtie
        find $dir -name *?gtf|grep -v '.*merge.gtf'|sort > $dir/mergelist.txt
        source activate RNA
        if [ ! -f $dir/stringtie_merge.gtf ];then
                stringtie --merge -p 8 -G $gene -o $dir/stringtie_merge.gtf $dir/mergelist.txt
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::转录本合并耗时%.2f分钟\n" $(echo `date`) $dur
        fi      
        source deactivate RNA
}

count(){
        printf "[%s %s %s %s %s %s]::转录本（基因）的定量：stringtie\n" $(echo `date`)  
        start=$(date +%s.%N)
        dir=$output/3_ballgown
        file_num=`ls -l $dir|wc -l`
        source activate RNA
        if [ $file_num -lt 3 ];then
                for id in $(cat $output/2_stringtie/mergelist.txt);do
                        base_name=$(basename $id)       
                        name=${base_name%.gtf}
                        read -u$1
                        {
                        stringtie -B -p 8 -G $output/2_stringtie/stringtie_merge.gtf -o $dir/$name/$base_name $output/1_hisat/${name}.bam &>> $dir/count.log
                        printf "[%s %s %s %s %s %s]::%s转录本定量完成\n" $(echo `date`) $name
                        echo >&$1
                        } &
                done && wait
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::转录本合并耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}


settings
index
thread 4 mapping
thread 6 assemble
merge
thread 6 count

差异表达分析ballgown

设置工作目录

就设置为当前目录下，新建一个DE目录，根据需求调整

!pwd
!mkdir DE

/root/python_code_scq/python_bio_file
mkdir: 无法创建目录"DE": 文件已存在

setwd("/root/python_code_scq/python_bio_file/DE")
suppressMessages(library(ballgown))
suppressMessages(library(genefilter))
suppressMessages(library(dplyr))

读取样本的表型数据

pheno_data=read.csv("/root/python_code_scq/python_bio_file/geuvadis_phenodata.csv")
pheno_data
idssexpopulation
ERR188044male YRI
ERR188104male YRI
ERR188234female YRI
ERR188245female GBR
ERR188257male GBR
ERR188273female YRI
ERR188337female GBR
ERR188383male GBR
ERR188401male GBR
ERR188428female GBR
ERR188454male YRI
ERR204916female YRI

读取定量分析的表达数据

bg_chrX = ballgown(dataDir = "../3_ballgown",samplePattern = "ERR", pData = pheno_data )

Sat Dec 29 10:56:05 2018
Sat Dec 29 10:56:05 2018: Reading linking tables
Sat Dec 29 10:56:06 2018: Reading intron data files
Sat Dec 29 10:56:07 2018: Merging intron data
Sat Dec 29 10:56:07 2018: Reading exon data files
Sat Dec 29 10:56:09 2018: Merging exon data
Sat Dec 29 10:56:10 2018: Reading transcript data files
Sat Dec 29 10:56:10 2018: Merging transcript data
Wrapping up the results
Sat Dec 29 10:56:10 2018

bg_chrX

ballgown instance with 3354 transcripts and 12 samples

过滤掉表达量低的基因

bg_chrX_filt=subset(bg_chrX,"rowVars(texpr(bg_chrX))>1",genomesubset=TRUE)

对转录本进行差异分析

results_transcripts=stattest(bg_chrX_filt,feature="transcript",covariate="sex",adjustvars=c("population"),getFC=TRUE,meas="FPKM")

对基因进行差异分析

results_genes=stattest(bg_chrX_filt,feature="gene",covariate="sex",adjustvars=c("population"),getFC=TRUE,meas="FPKM")

给转录本添加基因名和基因ID

results_transcripts=data.frame(geneNames=ballgown::geneNames(bg_chrX_filt),geneIDs=geneIDs(bg_chrX_filt),results_transcripts)

根据p-value值对分析结果进行排序(从小到大)

results_transcripts=arrange(results_transcripts,pval)
results_genes=arrange(results_genes,pval)
write.csv(results_transcripts, "chrX_transcript_results.csv",row.names=FALSE)
write.csv(results_genes, "chrX_gene_results.csv",row.names=FALSE)

提取显著差异表达的基因和转录本（q-value<0.05）

results_transcripts_0.05=subset(results_transcripts,results_transcripts$qval<0.05)
results_genes_0.05=subset(results_genes,results_genes$qval<0.05)
write.csv(results_transcripts_0.05,file="chrX_transcript_0.05.csv",row.names=FALSE)
write.csv(results_genes_0.05,file="chrX_genes_0.05.csv",row.names=FALSE)

数据可视化之颜色设定

tropical= c('darkorange', 'dodgerblue','hotpink', 'limegreen', 'yellow')
palette(tropical)
#当然rainbow()函数也可以完成这个任务
#palette(rainbow(5))

以FPKM为参考值作图，以性别作为区分条件

options(repr.plot.width = 9, repr.plot.height = 6)
fpkm = texpr(bg_chrX,meas="FPKM")
#方便作图将其log转换，+1是为了避免出现log2(0)的情况
fpkm = log2(fpkm+1)
# tiff(filename ="FPKM.tiff", compression = "lzw")
boxplot(fpkm,col=as.numeric(pheno_data$sex),las=2,ylab='log2(FPKM+1)')
# dev.off()

在这里插入图片描述

suppressMessages(library(reshape))
suppressMessages(library(ggplot2))
fpkm_L <- melt(fpkm)
tail(fpkm_L)
X1X2value
402433349 FPKM.ERR2049160.00000000
402443350 FPKM.ERR2049163.18196839
402453351 FPKM.ERR2049160.04649665
402463352 FPKM.ERR2049160.00000000
402473353 FPKM.ERR2049160.00000000
402483354 FPKM.ERR2049160.00000000 
colnames(fpkm_L)=c('n','sample','value')
head(fpkm_L)
nsamplevalue
1 FPKM.ERR1880442.218154
2 FPKM.ERR1880440.000000
3 FPKM.ERR1880445.333466
4 FPKM.ERR1880440.000000
5 FPKM.ERR1880440.000000
6 FPKM.ERR1880448.949343 
group_list=as.character(pheno_data$sex)
group_list
  1. 'male'
  2. 'male'
  3. 'female'
  4. 'female'
  5. 'male'
  6. 'female'
  7. 'female'
  8. 'male'
  9. 'male'
  10. 'female'
  11. 'male'
  12. 'female'
fpkm_L$group=rep(group_list,each=nrow(fpkm))
tail(fpkm_L)
nsamplevaluegroup
402433349 FPKM.ERR2049160.00000000 female
402443350 FPKM.ERR2049163.18196839 female
402453351 FPKM.ERR2049160.04649665 female
402463352 FPKM.ERR2049160.00000000 female
402473353 FPKM.ERR2049160.00000000 female
402483354 FPKM.ERR2049160.00000000 female 
options(repr.plot.width = 9, repr.plot.height = 5)
p=ggplot(fpkm_L,aes(x=sample,y=value,fill=group))+
geom_boxplot() + 
theme(text=element_text(face='bold'),axis.text.x=element_text(angle=30,hjust=1,size = 6))
p

在这里插入图片描述

#?graphics::boxplot

Subread -> featureCounts -> DESeq2

流程代码

#!/usr/bin/env bash

set -ue

settings(){
        samples=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/samples
        index=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genome/index
    output=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/subread+featurecounts
    if test -w $(dirname $output) &&  test -w $(dirname index);then
        mkdir -p {$index/subread,$output/1_subjunc,$output/2_featurecounts}   
    else
        echo "没有写入权限"
    fi
        genome=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genome/chrX.fa
    gene=/data/Data_base/test_tmp/RNA_seq_practice/chrX_data/genes/chr_X.gtf
        
}

thread(){
    tmp_fifofile="/tmp/$$.fifo" 
    mkfifo "$tmp_fifofile" 
    exec 6<>"$tmp_fifofile"  
    rm $tmp_fifofile
    thread_num=$1 
    for((i=0;i<$thread_num;i++));do 
        echo 
    done >&6 
    $2 6 
    exec 6>&- 
}

index(){
    printf "[%s %s %s %s %s %s]::建立索引subread-buildindex\n" $(echo `date`)
    start=$(date +%s.%N)
    file_num=`ls -l $index/subread|wc -l`   
    source activate RNA
        base_name=$(basename $genome)
        name=`awk -v each=$base_name 'BEGIN{split(each,arr,".");print arr[1]}' ` 
    if [ $file_num -lt 2 ];then
        subread-buildindex -o $index/subread/$name $genome &> $index/subread/index.log
    fi
    dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
    printf "[%s %s %s %s %s %s]::建立索引耗时%.2f分钟\n" $(echo `date`) $dur
    source deactivate RNA
}



mapping(){
        printf "[%s %s %s %s %s %s]::与参考基因组比对subjunc\n" $(echo `date`)
        start=$(date +%s.%N)
        dir=$output/1_subjunc
        find $samples/cleandata -name *1?f*q.gz|sort > $dir/1
        find $samples/cleandata -name *2?f*q.gz|sort > $dir/2
        paste -d ":" $dir/1 $dir/2  > $dir/config && rm $dir/1 $dir/2
        file_num=`ls -l $dir|wc -l`
        index_prefix=`awk -v each=$(basename $genome) 'BEGIN{split(each,arr,".");print arr[1]}' `
        source activate RNA
        if [ $file_num -lt 3 ];then
                for id in $(cat $dir/config);do
                        fq1=$(echo $id|cut -d":" -f1)
                        fq2=$(echo $id |cut -d":" -f2)
                        name=`awk -v each=$(basename $fq1) 'BEGIN{split(each,arr,"_");print arr[1]}' ` 
                        read -u$1
                        {
                        subjunc -T 5 -i $index/subread/$index_prefix -r $fq1 -R $fq2 -o $dir/${name}.subjunc.bam &> $dir/${name}.log
                        printf "[%s %s %s %s %s %s]::%s比对完成\n" $(echo `date`) $name
                        echo >&$1       
                        } &
                done && wait
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::比对耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA

}


count(){
        printf "[%s %s %s %s %s %s]:: 转录本定量featureCounts\n" $(echo `date`)         
        start=$(date +%s.%N)
        dir=$output/2_featurecounts
        file_num=`ls -l $dir|wc -l`
        source activate RNA
        if [ $file_num -lt 3 ];then
                featureCounts -T 5 -p -t exon -g gene_id  -a $gene -o $dir/all.id.txt  $output/1_subjunc/*.bam  1>$dir/counts.log 2>&1 
                # -g 默认gene_id, 可以选择gene_name，后面可以用R将ID转换成gene symbol
                # 这样得到的all.id.txt文件就是表达矩阵，但是，这个featureCounts有非常多的参数可以调整
                dur=$(echo "($(date +%s.%N) - $start)/60" | bc -l)
                printf "[%s %s %s %s %s %s]::转录本定量耗时%.2f分钟\n" $(echo `date`) $dur
        fi
        source deactivate RNA
}

settings
index
thread 3 mapping
count

DESeq2差异分析

差异分析需要的包：

  • limma
  • edgeR
  • gplots
  • org.Mm.eg.db
  • RColorBrewer
  • Glimma
  • DESeq2

DESeq2差异分析的三个主要数据：

  • 表达矩阵（原始的count数据）
  • 分组矩阵（实验设计的分组情况）
  • 差异比较矩阵（指定差异比较对象）

DESeq2差异分析的三个主要步骤：

  • 构建DESeqDataSet（dds对象）
  • DESeq()函数进行差异分析
  • results()函数提取差异分析结果

构建DESeqDataSet对象有两种方式，一种是通过DESeqDataSet()函数将SummarizedExperiment->DESeqDataSet,如果我们只有表达矩阵和样本表型数据，如何构建DESeqDataSet数据集,可以通过另一种方式，使用DESeqDataSetFromMatrix()函数将表达矩阵和样本表型数据转化为DESeqDataSet数据对象。

colData里面的信息，就相当于我们自己准备的表型数据，建议是csv，tsv格式（comma-separated value；tab- separated value）文件，可以把样本表型的数据赋值给colData对象

suppressMessages(library(edgeR))
suppressMessages(library(limma))
suppressMessages(library(Glimma))
suppressMessages(library(gplots))
suppressMessages(library(DESeq2))
suppressMessages(library(org.Hs.eg.db))
suppressMessages(library(RColorBrewer))

读取数据,提取表达矩阵

#读取count table数据
data <- read.table("all.id.txt",stringsAsFactors = F, header = T)
meta = data[,1:6]
exprSet = data[,7:ncol(data)]
rownames(exprSet) <- data[,1]
colnames(exprSet)
  1. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188044.subjunc.bam'
  2. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188104.subjunc.bam'
  3. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188234.subjunc.bam'
  4. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188245.subjunc.bam'
  5. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188257.subjunc.bam'
  6. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188273.subjunc.bam'
  7. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188337.subjunc.bam'
  8. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188383.subjunc.bam'
  9. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188401.subjunc.bam'
  10. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188428.subjunc.bam'
  11. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR188454.subjunc.bam'
  12. 'X.data.Data_base.test_tmp.RNA_seq_practice.chrX_data.subread.featurecounts.1_subjunc.ERR204916.subjunc.bam'
class(exprSet)

‘data.frame’

修改列名

# suppressMessages(library(dplyr))
# name <- lapply(strsplit(as.character(colnames(exprSet)),'[.]'),function(x){x[length(x)-2]}) %>% unlist
name <- unlist(lapply(strsplit(as.character(colnames(exprSet)),'[.]'),function(x){x[length(x)-2]}))
colnames(exprSet) <- name 
colnames(exprSet)
  1. 'ERR188044'
  2. 'ERR188104'
  3. 'ERR188234'
  4. 'ERR188245'
  5. 'ERR188257'
  6. 'ERR188273'
  7. 'ERR188337'
  8. 'ERR188383'
  9. 'ERR188401'
  10. 'ERR188428'
  11. 'ERR188454'
  12. 'ERR204916'
head(exprSet,6)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000228572 1 0 0 0 0 0 1 0 0 0 1 0
ENSG000001823784128264631381631100720593251 8661944897 15172587
ENSG000001786051235 854140610351437 343179212461967606 12481469
ENSG00000226179 13 13 9 4 1 2 6 3 8 4 7 10
ENSG00000167393 243 173 236 282 171 76 279 60 327111 96 258
ENSG00000275287 0 1 0 0 0 0 0 0 0 0 0 0
 
exprSet[rownames(exprSet)=="ENSG00000205755",]
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG000002057559 105 3 4 2 8 4 6 3 2 5

查看数据维度，即几行几列

dim(exprSet)
  1. 2450
  2. 12

表型数据读取,样本分组信息

# Read the sample information into R
coldata <- read.csv("geuvadis_phenodata.csv")
coldata
idssexpopulation
ERR188044male YRI
ERR188104male YRI
ERR188234female YRI
ERR188245female GBR
ERR188257male GBR
ERR188273female YRI
ERR188337female GBR
ERR188383male GBR
ERR188401male GBR
ERR188428female GBR
ERR188454male YRI
ERR204916female YRI 
coldata$group <- factor(paste0(coldata$sex, coldata$population))

coldata
idssexpopulationgroup
ERR188044male YRI maleYRI
ERR188104male YRI maleYRI
ERR188234female YRI femaleYRI
ERR188245female GBR femaleGBR
ERR188257male GBR maleGBR
ERR188273female YRI femaleYRI
ERR188337female GBR femaleGBR
ERR188383male GBR maleGBR
ERR188401male GBR maleGBR
ERR188428female GBR femaleGBR
ERR188454male YRI maleYRI
ERR204916female YRI femaleYRI
 
suppressMessages(library(pheatmap))
suppressMessages(library(corrplot))

可视化样本间的相似性

options(repr.plot.width = 5, repr.plot.height = 5)
# png('cor.png')
corrplot(cor(log2(exprSet+1)))
# dev.off()

在这里插入图片描述

m=cor(log2(exprSet+1))
#cor矩阵用scale归一化
pheatmap(scale(cor(log2(exprSet+1))))

构建DESeqDataSet（dds）对象

#将读取数据由data.frame转换成matrix类型
countdata <- as.matrix(exprSet)
#将countdata转换成DESeq2的数据格式（构建dds）
dds <- DESeqDataSetFromMatrix(countData = countdata,colData = coldata, design = ~ group )
dds

class: DESeqDataSet 
dim: 2450 12 
metadata(1): version
assays(1): counts
rownames(2450): ENSG00000228572 ENSG00000182378 ... ENSG00000276543
  ENSG00000227159
rowData names(0):
colnames(12): ERR188044 ERR188104 ... ERR188454 ERR204916
colData names(4): ids sex population group

#将所有样本中count值均为0的基因所在行去除
dds <- dds[rowSums(counts(dds) > 0) != 0 ,] #dds[rowSums(assay(dds) > 0) != 0 , ]
dds

class: DESeqDataSet 
dim: 1412 12 
metadata(1): version
assays(1): counts
rownames(1412): ENSG00000228572 ENSG00000182378 ... ENSG00000182484
  ENSG00000227159
rowData names(0):
colnames(12): ERR188044 ERR188104 ... ERR188454 ERR204916
colData names(4): ids sex population group

# DESeq分析，大小因子的估计，离差估计，负二项分布的拟合以及计算相应的统计量
dds <- DESeq(dds, parallel = T) #parallel = T程序并行，提高速度

estimating size factors
estimating dispersions
gene-wise dispersion estimates: 1 workers
mean-dispersion relationship
final dispersion estimates, fitting model and testing: 1 workers

plotDispEsts(dds, main="Dispersion plot")

在这里插入图片描述

resultsNames(dds)
  1. 'Intercept'
  2. 'group_femaleYRI_vs_femaleGBR'
  3. 'group_maleGBR_vs_femaleGBR'
  4. 'group_maleYRI_vs_femaleGBR'

使用rlog-transformed归一化数据

rld <- rlog(dds)
countdata_rlog <- assay(rld) 
head(assay(rld))
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000228572-1.716339-1.743507-1.743821-1.737338-1.739732-1.731162-1.716953-1.740015-1.744537-1.736857-1.713780-1.740743
ENSG0000018237811.46140810.98992611.13860410.96376810.32728511.67554811.16820710.16994810.61162710.41876510.64481511.174434
ENSG0000017860510.007361 9.64054010.09599610.26933310.421237 9.64724210.33147510.25630310.378303 9.78266910.20476410.367591
ENSG00000226179 2.753014 2.747507 2.620197 2.538756 2.352026 2.503809 2.517179 2.450266 2.573627 2.546540 2.612662 2.723157
ENSG00000167393 7.540045 7.216756 7.476484 8.109476 7.458899 7.303154 7.627266 6.576983 7.734151 7.262772 6.894270 7.775687
ENSG00000275287-2.216337-2.207251-2.216517-2.214314-2.215128-2.212215-2.216546-2.215224-2.216760-2.214150-2.215468-2.215471

使用rlog-transformed 与 正常 区别

options(repr.plot.width = 9, repr.plot.height = 9)
par(cex = 0.7)
n.sample=ncol(countdata)
if(n.sample>40) par(cex = 0.5)
cols <- rainbow(n.sample*1.2)
par(mfrow=c(2,2))
boxplot(countdata, col = cols,main="expression value",las=2)
boxplot(countdata_rlog, col = cols,main="expression value",las=2)
hist(countdata)
hist(countdata_rlog)

在这里插入图片描述

热图呈现样品间的距离

mycols <- brewer.pal(8, "Dark2")[1:length(unique(coldata$group))]
sampleDists <- as.matrix(dist(t(countdata_rlog)))
heatmap.2(as.matrix(sampleDists), key=F, trace="none",
            col=colorpanel(100, "black", "white"),
            ColSideColors=mycols[coldata$group], RowSideColors=mycols[coldata$group],
margin=c(10, 10), main="Sample Distance Matrix")

在这里插入图片描述

差异分析及结果提取

#提取性别间差异分析的结果
sex_Y_result <- results(dds, contrast = c("group","maleYRI","femaleYRI"), parallel = T)
#?results 查看用法
sex_G_result <- results(dds, contrast = c("group","maleGBR","femaleGBR"), parallel = T)
sex_G_result

log2 fold change (MLE): group maleGBR vs femaleGBR 
Wald test p-value: group maleGBR vs femaleGBR 
DataFrame with 1412 rows and 6 columns
                         baseMean     log2FoldChange             lfcSE
                        <numeric>          <numeric>         <numeric>
ENSG00000228572 0.212064658186015  -1.18134449838418  4.40721523394822
ENSG00000182378   2080.8358537599 -0.759383223494156 0.369671551206202
ENSG00000178605  1146.43865301777  0.288751709902595 0.319524288527945
ENSG00000226179  6.08058463520939 -0.493246072149153 0.820198203867892
ENSG00000167393  183.376192726023 -0.562850597829534 0.484234566315877
...                           ...                ...               ...
ENSG00000124334  60.2368766837945  0.489304618353469 0.581968571149684
ENSG00000270726 0.209832834505338  -0.22070381980218  4.40735632395141
ENSG00000185203  1.03705472979752   1.36980197787972  2.33437609518548
ENSG00000182484  493.281659491792  0.264865107532117 0.359069049586416
ENSG00000227159  3.18504610277688 -0.645544611222065  1.25294052192237
                               stat             pvalue              padj
                          <numeric>          <numeric>         <numeric>
ENSG00000228572  -0.268047834216135  0.788662499091892 0.992519295917494
ENSG00000182378   -2.05421061214032 0.0399553117393148 0.992519295917494
ENSG00000178605   0.903692521256771  0.366158465847581 0.992519295917494
ENSG00000226179  -0.601374240790949  0.547590751585147 0.992519295917494
ENSG00000167393   -1.16235113513638  0.245092863468186 0.992519295917494
...                             ...                ...               ...
ENSG00000124334   0.840774987877512  0.400474001859852 0.992519295917494
ENSG00000270726 -0.0500762369955846  0.960061636105264 0.992519295917494
ENSG00000185203    0.58679575270877  0.557340889682515 0.992519295917494
ENSG00000182484   0.737643937390858  0.460730848276013 0.992519295917494
ENSG00000227159  -0.515223667785613  0.606396732014109 0.992519295917494

结果说明

  • 第一列基因名
  • baseMean 经过矫正的reads count均值
  • log2FoldChange 对差异倍数取以2为底的对数
  • lfcSE 差异倍数取对数后的标准差
  • stat Wald统计量，由log2FoldChange除以标准差所得
  • pvalue 原始的p值
  • padj 校正后的p值
summary(sex_G_result)

out of 1412 with nonzero total read count
adjusted p-value < 0.1
LFC > 0 (up)       : 2, 0.14%
LFC < 0 (down)     : 13, 0.92%
outliers [1]       : 1, 0.071%
low counts [2]     : 0, 0%
(mean count < 0)
[1] see 'cooksCutoff' argument of ?results
[2] see 'independentFiltering' argument of ?results

#在群体间差异分析的基因分析结果
population_F_result <- results(dds, contrast = c("group","femaleYRI","femaleGBR"), parallel = T)
population_M_result <- results(dds, contrast = c("group","maleYRI","maleGBR"), parallel = T)

#提取性别间差异表达的基因，并存入文件
sex_Y_result_0.01 <- sex_Y_result[which(sex_Y_result$pvalue < 0.01),]
write.csv(sex_Y_result_0.01, file="DE/sex_Y_result_0.01_DESeq2.csv")

#提取种群间差异表达的基因，并存入文件
population_F_result_0.01 <- subset(population_F_result,pvalue < 0.01)
write.csv(population_F_result_0.01, file="DE/population_F_result_0.01_DESeq2.csv")

流程参考文献RNA-Seq workflow: gene-level exploratory analysis and differential expression

edgeR差异分析

edgeR差异分析的三个主要数据：

  • 表达矩阵（原始的count数据）
  • 分组矩阵（实验设计的分组情况）
  • 差异比较矩阵（指定差异比较对象）

edgeR差异分析的五个主要步骤：

  • DGElist()函数构造DGEList对象，转化counts 成DGEList对象
  • model.matrix()函数构建设计矩阵
  • estimateDisp()函数进行差异分析
  • glmQLFit()函数构建DGEGLM对象
  • glmQLFTest()函数进行QL F-test

读取数据,提取表达矩阵

上面已经做过了

表型数据读取,样本分组信息

上面已经做过了

#再转换为数据框
class(countdata)
head(countdata,3)
countdata <- as.data.frame(countdata)
dim(countdata)
class(countdata)

‘matrix’

ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000228572 1 0 0 0 0 0 1 0 0 0 1 0
ENSG000001823784128264631381631100720593251 8661944897 15172587
ENSG000001786051235 854140610351437 343179212461967606 12481469
  1. 2450
  2. 12

‘data.frame’

coldata
idssexpopulationgroup
ERR188044male YRI maleYRI
ERR188104male YRI maleYRI
ERR188234female YRI femaleYRI
ERR188245female GBR femaleGBR
ERR188257male GBR maleGBR
ERR188273female YRI femaleYRI
ERR188337female GBR femaleGBR
ERR188383male GBR maleGBR
ERR188401male GBR maleGBR
ERR188428female GBR femaleGBR
ERR188454male YRI maleYRI
ERR204916female YRI femaleYRI

过滤表达量低的基因

在这个数据集中，我们选择保留至少在两个样本中的CPM（counts-per-million）值大于0.5的基因(根据实际情况作调整)

使用edgeR 包中的cpm 函数

myCPM <- cpm(countdata)
# Have a look at the output
head(myCPM)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000228572 0.8414109 0.0000000 0.000000 0.00000 0.000000 0.000000 0.9116842 0.000000 0.0000 0.00000 1.058151 0.00000
ENSG000001823783473.34410342263.37430962255.726656 2125.22542 1185.821950 3899.517816 2963.8854523 976.066093 1629.1449 1167.61147 1605.214567 2623.31454
ENSG000001786051039.1424340 730.50705231010.692058 1348.62557 1692.180876 649.603988 1633.73815151404.362993 1648.4198 788.82113 1320.572036 1489.62082
ENSG00000226179 10.9383414 11.1201308 6.469579 5.21208 1.177579 3.787778 5.4701054 3.381291 6.7043 5.20674 7.407055 10.14037
ENSG00000167393 204.4628433 147.9832787 169.646747 367.45161 201.365992 143.935577 254.3599019 67.625826 274.0383 144.48704 101.582464 261.62163
ENSG00000275287 0.0000000 0.8553947 0.000000 0.00000 0.000000 0.000000 0.0000000 0.000000 0.0000 0.00000 0.000000 0.00000 
# Which values in myCPM are greater than 0.5?
thresh <- myCPM > 0.5
# This produces a logical matrix with TRUEs and FALSEs
head(thresh)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000228572 TRUEFALSEFALSEFALSEFALSEFALSE TRUEFALSEFALSEFALSE TRUEFALSE
ENSG00000182378 TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
ENSG00000178605 TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
ENSG00000226179 TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
ENSG00000167393 TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
ENSG00000275287FALSE TRUEFALSEFALSEFALSEFALSEFALSEFALSEFALSEFALSEFALSEFALSE
 
# Summary of how many TRUEs there are in each row
# There are 672  genes that have TRUEs in all 12 samples.
table(rowSums(thresh))

   0    1    2    3    4    5    6    7    8    9   10   11   12 
1038  208  106   79   58   45   40   44   28   34   40   58  672 

# we would like to keep genes that have at least 2 TRUES in each row of thresh
keep <- rowSums(thresh) >= 2
# Subset the rows of countdata to keep the more highly expressed genes
counts.keep <- countdata[keep,]
summary(keep)

   Mode   FALSE    TRUE 
logical    1246    1204 

dim(counts.keep)
  1. 1204
  2. 12

ID转换，基因注释

# rm(list=ls())
ls('package:org.Hs.eg.db')
  1. 'org.Hs.eg'
  2. 'org.Hs.eg_dbconn'
  3. 'org.Hs.eg_dbfile'
  4. 'org.Hs.eg_dbInfo'
  5. 'org.Hs.eg_dbschema'
  6. 'org.Hs.eg.db'
  7. 'org.Hs.egACCNUM'
  8. 'org.Hs.egACCNUM2EG'
  9. 'org.Hs.egALIAS2EG'
  10. 'org.Hs.egCHR'
  11. 'org.Hs.egCHRLENGTHS'
  12. 'org.Hs.egCHRLOC'
  13. 'org.Hs.egCHRLOCEND'
  14. 'org.Hs.egENSEMBL'
  15. 'org.Hs.egENSEMBL2EG'
  16. 'org.Hs.egENSEMBLPROT'
  17. 'org.Hs.egENSEMBLPROT2EG'
  18. 'org.Hs.egENSEMBLTRANS'
  19. 'org.Hs.egENSEMBLTRANS2EG'
  20. 'org.Hs.egENZYME'
  21. 'org.Hs.egENZYME2EG'
  22. 'org.Hs.egGENENAME'
  23. 'org.Hs.egGO'
  24. 'org.Hs.egGO2ALLEGS'
  25. 'org.Hs.egGO2EG'
  26. 'org.Hs.egMAP'
  27. 'org.Hs.egMAP2EG'
  28. 'org.Hs.egMAPCOUNTS'
  29. 'org.Hs.egOMIM'
  30. 'org.Hs.egOMIM2EG'
  31. 'org.Hs.egORGANISM'
  32. 'org.Hs.egPATH'
  33. 'org.Hs.egPATH2EG'
  34. 'org.Hs.egPFAM'
  35. 'org.Hs.egPMID'
  36. 'org.Hs.egPMID2EG'
  37. 'org.Hs.egPROSITE'
  38. 'org.Hs.egREFSEQ'
  39. 'org.Hs.egREFSEQ2EG'
  40. 'org.Hs.egSYMBOL'
  41. 'org.Hs.egSYMBOL2EG'
  42. 'org.Hs.egUCSCKG'
  43. 'org.Hs.egUNIGENE'
  44. 'org.Hs.egUNIGENE2EG'
  45. 'org.Hs.egUNIPROT'
entrezID2ensemblID <- toTable(org.Hs.egENSEMBL)
head(entrezID2ensemblID)
gene_idensembl_id
1 ENSG00000121410
2 ENSG00000175899
3 ENSG00000256069
9 ENSG00000171428
10 ENSG00000156006
12 ENSG00000196136
 
entrezID2symbol <- toTable(org.Hs.egSYMBOL)
head(entrezID2symbol)
gene_idsymbol
1 A1BG
2 A2M
3 A2MP1
9 NAT1
10 NAT2
11 NATP 
columns(org.Hs.eg.db)
  1. 'ACCNUM'
  2. 'ALIAS'
  3. 'ENSEMBL'
  4. 'ENSEMBLPROT'
  5. 'ENSEMBLTRANS'
  6. 'ENTREZID'
  7. 'ENZYME'
  8. 'EVIDENCE'
  9. 'EVIDENCEALL'
  10. 'GENENAME'
  11. 'GO'
  12. 'GOALL'
  13. 'IPI'
  14. 'MAP'
  15. 'OMIM'
  16. 'ONTOLOGY'
  17. 'ONTOLOGYALL'
  18. 'PATH'
  19. 'PFAM'
  20. 'PMID'
  21. 'PROSITE'
  22. 'REFSEQ'
  23. 'SYMBOL'
  24. 'UCSCKG'
  25. 'UNIGENE'
  26. 'UNIPROT'
help('select')

ann <- select(org.Hs.eg.db,keys = rownames(counts.keep), keytype="ENSEMBL",columns = c("ENSEMBL","SYMBOL","GENENAME"))

'select()' returned 1:many mapping between keys and columns

head(ann)
dim(ann)
ENSEMBLSYMBOLGENENAME
ENSG00000228572 NA NA
ENSG00000182378 PLCXD1 phosphatidylinositol specific phospholipase C X domain containing 1
ENSG00000178605 GTPBP6 GTP binding protein 6 (putative)
ENSG00000226179 NA NA
ENSG00000167393 PPP2R3B protein phosphatase 2 regulatory subunit B''beta
ENSG00000237531 NA NA
  1. 1213
  2. 3
symbol <-  ann[match(rownames(counts.keep),ann$ENSEMBL),2]
rownames(counts.keep) <- paste(rownames(counts.keep) ,':', symbol)
head(counts.keep)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000228572 : NA 1 0 0 0 0 0 1 0 0 0 1 0
ENSG00000182378 : PLCXD14128264631381631100720593251 8661944897 15172587
ENSG00000178605 : GTPBP61235 854140610351437 343179212461967606 12481469
ENSG00000226179 : NA 13 13 9 4 1 2 6 3 8 4 7 10
ENSG00000167393 : PPP2R3B 243 173 236 282 171 76 279 60 327111 96 258
ENSG00000237531 : NA 0 1 4 0 0 0 4 0 0 6 2 0
 
counts.keep <- counts.keep[grep(".*: NA$",rownames(counts.keep),invert = T), ]
head(counts.keep)
dim(counts.keep)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000182378 : PLCXD14128264631381631100720593251 8661944897 15172587
ENSG00000178605 : GTPBP61235 854140610351437 343179212461967606 12481469
ENSG00000167393 : PPP2R3B 243 173 236 282 171 76 279 60 327111 96 258
ENSG00000205755 : CRLF2 9 10 5 3 4 2 8 4 6 3 2 5
ENSG00000198223 : CSF2RA 7 2 3 1 1 2 1 6 2 0 0 7
ENSG00000185291 : IL3RA 651 664 195 202 455 216 283 413 693221 593 620
  1. 697
  2. 12

构建DGEList对象（edgeR的数据分析格式）

edgeR_data <- DGEList(counts = counts.keep, group = coldata$group)
# See what slots are stored in edgeR_data
names(edgeR_data)
head(edgeR_data$counts)
  1. 'counts'
  2. 'samples'
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000182378 : PLCXD14128264631381631100720593251 8661944897 15172587
ENSG00000178605 : GTPBP61235 854140610351437 343179212461967606 12481469
ENSG00000167393 : PPP2R3B 243 173 236 282 171 76 279 60 327111 96 258
ENSG00000205755 : CRLF2 9 10 5 3 4 2 8 4 6 3 2 5
ENSG00000198223 : CSF2RA 7 2 3 1 1 2 1 6 2 0 0 7
ENSG00000185291 : IL3RA 651 664 195 202 455 216 283 413 693221 593 620
 
# Library size information is stored in the samples slot
edgeR_data$samples
grouplib.sizenorm.factors
ERR188044maleYRI 1062212 1
ERR188104maleYRI 1037630 1
ERR188234femaleYRI1236643 1
ERR188245femaleGBR 688237 1
ERR188257maleGBR 769281 1
ERR188273femaleYRI 476704 1
ERR188337femaleGBR1009898 1
ERR188383maleGBR 786937 1
ERR188401maleGBR 1087875 1
ERR188428femaleGBR 697609 1
ERR188454maleYRI 836018 1
ERR204916femaleYRI 876941 1

可以自己手动计算一下，看一致否

apply(counts.keep,2,sum)
ERR188044
1062212
ERR188104
1037630
ERR188234
1236643
ERR188245
688237
ERR188257
769281
ERR188273
476704
ERR188337
1009898
ERR188383
786937
ERR188401
1087875
ERR188428
697609
ERR188454
836018
ERR204916
876941

质控

文库大小及其分布图

edgeR_data$samples$lib.size
  1. 1062212
  2. 1037630
  3. 1236643
  4. 688237
  5. 769281
  6. 476704
  7. 1009898
  8. 786937
  9. 1087875
  10. 697609
  11. 836018
  12. 876941
# The names argument tells the barplot to use the sample names on the x-axis
# The las argument rotates the axis names
options(repr.plot.width = 6, repr.plot.height = 4)
barplot(edgeR_data$samples$lib.size,names=colnames(edgeR_data),las=2)
title("Barplot of library sizes")

在这里插入图片描述

# Get log2 counts per million
logcounts <- cpm(edgeR_data,log=TRUE)
# Check distributions of samples using boxplots
boxplot(logcounts, xlab="", ylab="Log2 counts per million",las=2)
# Let's add a blue horizontal line that corresponds to the median logCPM
abline(h=median(logcounts),col="blue")
title("Boxplots of logCPMs (unnormalised)")

在这里插入图片描述

Multidimensional scaling plots

A principle components analysis is an example of an unsupervised analysis, where we don’t need to specify the groups.

plotMDS(edgeR_data)

在这里插入图片描述

按组添加颜色

# We specify the option to let us plot two plots side-by-sde
par(mfrow=c(1,2))
options(repr.plot.width = 9, repr.plot.height = 5)
# Let's set up colour schemes for sex
levels(coldata$sex)
## Let's choose purple for basal and orange for luminal
col.sex <- c("purple","orange")[coldata$sex]
data.frame(coldata$sex,col.sex)
# Redo the MDS with cell type colouring
plotMDS(edgeR_data,col=col.sex)
# Let's add a legend to the plot so we know which colours correspond to which sex
legend("topleft",fill=c("purple","orange"),legend=levels(coldata$sex))
# Add a title
title("sex")

# Similarly for population
levels(coldata$population)
col.population <- c("blue","red")[coldata$population]
plotMDS(edgeR_data,col=col.population)
legend("topleft",fill=c("blue","red"),legend=levels(coldata$population),cex=0.8)
title("population")
  1. 'female'
  2. 'male'
coldata.sexcol.sex
male orange
male orange
femalepurple
femalepurple
male orange
femalepurple
femalepurple
male orange
male orange
femalepurple
male orange
femalepurple
  1. 'GBR'
  2. 'YRI'

在这里插入图片描述

整合成一张图

cols <- c("#1B9E77", "#D95F02")
col.population <- c("#1B9E77", "#D95F02")[coldata$population]
char.sex <- c(1,4)[coldata$sex]
char.sex
  1. 4
  2. 4
  3. 1
  4. 1
  5. 4
  6. 1
  7. 1
  8. 4
  9. 4
  10. 1
  11. 4
  12. 1
options(repr.plot.width = 6, repr.plot.height = 5)
plotMDS(edgeR_data,dim=c(1,2),col=col.population,pch=char.sex,cex=2)
legend("top",legend=levels(coldata$population),col=cols,pch=16)
legend("right",legend=levels(coldata$sex),pch=c(1,4))

在这里插入图片描述

标准化，归一化

Normalisation for composition bias

# Apply normalisation to DGEList object
edgeR_data <- calcNormFactors(edgeR_data)

This will update the normalisation factors in the DGEList object (their default values are 1). Take a look at the normalisation factors for these samples.

edgeR_data$samples
grouplib.sizenorm.factors
ERR188044maleYRI 1062212 1.0091479
ERR188104maleYRI 1037630 1.0063310
ERR188234femaleYRI1236643 0.8810990
ERR188245femaleGBR 688237 0.9781766
ERR188257maleGBR 769281 1.0374696
ERR188273femaleYRI 476704 0.9219804
ERR188337femaleGBR1009898 1.0665011
ERR188383maleGBR 786937 1.0556198
ERR188401maleGBR 1087875 1.0580396
ERR188428femaleGBR 697609 0.9437640
ERR188454maleYRI 836018 1.0681350
ERR204916femaleYRI 876941 0.9947293

ERR188234样本的normalisation factors最小，ERR188337样本的normalisation factors最大，
If we plot mean difference plots using the plotMD function for these samples, we should be able to see the composition bias problem. We will use the logcounts, which have been normalised for library size, but not for composition bias.

options(repr.plot.width = 9, repr.plot.height = 5)
par(mfrow=c(1,2))
plotMD(logcounts,column = 3)
abline(h=0,col="grey")
plotMD(logcounts,column = 7)
abline(h=0,col="grey")

在这里插入图片描述

The mean-difference plots show average expression (mean: x-axis) against log-fold-changes (difference: y-axis). Because our DGEList object contains the normalisation factors, if we redo these plots using edgeR_data, we should see the composition bias problem has been solved.

par(mfrow=c(1,2))
plotMD(edgeR_data,column = 3)
abline(h=0,col="grey")
plotMD(edgeR_data,column = 7)
abline(h=0,col="grey")

在这里插入图片描述

热图层次聚类

# We estimate the variance for each row in the logcounts matrix
var_genes <- apply(logcounts, 1, var)
head(var_genes)
ENSG00000182378 : PLCXD1
0.411283212250483
ENSG00000178605 : GTPBP6
0.233040789399784
ENSG00000167393 : PPP2R3B
0.440930844051291
ENSG00000205755 : CRLF2
0.267770197328411
ENSG00000198223 : CSF2RA
2.65972252131491
ENSG00000185291 : IL3RA
0.468356036950437
# Get the gene names for the top 50 most variable genes
select_var <- names(sort(var_genes, decreasing=TRUE))[1:50]
head(select_var)
  1. 'ENSG00000229807 : XIST'
  2. 'ENSG00000198759 : EGFL6'
  3. 'ENSG00000146938 : NLGN4X'
  4. 'ENSG00000269096 : CT45A3'
  5. 'ENSG00000175556 : LONRF3'
  6. 'ENSG00000003096 : KLHL13'
# Subset logcounts matrix
highly_variable_lcpm <- logcounts[select_var,]
dim(highly_variable_lcpm)
  1. 50
  2. 12
options(repr.plot.width = 9, repr.plot.height = 6)

my_heatmap <- function(countdata,col_label,row_lable,group){
    mypalette <- brewer.pal(11,"RdYlBu")
    morecols <- colorRampPalette(mypalette)
    col.sex <- c("purple","orange")[group]
    # Plot the heatmap
    heatmap.2(countdata,labCol=col_label,srtCol=15, labRow=row_lable,
              col=rev(morecols(50)),trace="none", main="Top 50 most variable genes across samples",
              ColSideColors=col.sex,colCol=col.sex,scale="row")
}


col_label <- paste(coldata$sex ,'.', coldata$population)
row_lable <- unlist(lapply(strsplit(rownames(highly_variable_lcpm),":"),function(x) x[2]))
group <- coldata$sex
my_heatmap(highly_variable_lcpm, col_label, row_lable, group)

在这里插入图片描述

建立设计矩阵

design <- model.matrix(~0 + group , data = coldata)
design
groupfemaleGBRgroupfemaleYRIgroupmaleGBRgroupmaleYRI
10001
20001
30100
41000
50010
60100
71000
80010
90010
101000
110001
120100
 
#修改设计矩阵的列名,行名
colnames(design)
  1. 'groupfemaleGBR'
  2. 'groupfemaleYRI'
  3. 'groupmaleGBR'
  4. 'groupmaleYRI'
levels(coldata$group)
  1. 'femaleGBR'
  2. 'femaleYRI'
  3. 'maleGBR'
  4. 'maleYRI'
colnames(design) <- levels(coldata$group)
colnames(design)
rownames(design) <- coldata$id
rownames(design)
design
  1. 'femaleGBR'
  2. 'femaleYRI'
  3. 'maleGBR'
  4. 'maleYRI'
  1. 'ERR188044'
  2. 'ERR188104'
  3. 'ERR188234'
  4. 'ERR188245'
  5. 'ERR188257'
  6. 'ERR188273'
  7. 'ERR188337'
  8. 'ERR188383'
  9. 'ERR188401'
  10. 'ERR188428'
  11. 'ERR188454'
  12. 'ERR204916'
femaleGBRfemaleYRImaleGBRmaleYRI
ERR1880440001
ERR1881040001
ERR1882340100
ERR1882451000
ERR1882570010
ERR1882730100
ERR1883371000
ERR1883830010
ERR1884010010
ERR1884281000
ERR1884540001
ERR2049160100

估计离散度

edgeR_data <- estimateDisp(edgeR_data, design)
names(edgeR_data)
  1. 'counts'
  2. 'samples'
  3. 'design'
  4. 'common.dispersion'
  5. 'trended.dispersion'
  6. 'tagwise.dispersion'
  7. 'AveLogCPM'
  8. 'trend.method'
  9. 'prior.df'
  10. 'prior.n'
  11. 'span'

构建DGEGLM对象

fit <- glmQLFit(edgeR_data, design)

建立对照

?makeContrasts

my.contrasts <- makeContrasts(YRI.f.vs.m = femaleYRI - maleYRI,  GBR.f.vs.m = femaleGBR - maleGBR, f.YRI.vs.GBR = femaleYRI - femaleGBR, m.YRI.vs.GBR = maleYRI - maleGBR, levels = design)

差异分析（用quasi-likelihood(QL) F-test）

qlf_YRI.f.vs.m <- glmQLFTest(fit, contrast = my.contrasts[, "YRI.f.vs.m"])
qlf_GBR.f.vs.m <- glmQLFTest(fit, contrast = my.contrasts[, "GBR.f.vs.m"])
qlf_f.YRI.vs.GBR <- glmQLFTest(fit, contrast = my.contrasts[, "f.YRI.vs.GBR"])
qlf_m.YRI.vs.GBR <- glmQLFTest(fit, contrast = my.contrasts[, "m.YRI.vs.GBR"])
names(qlf_m.YRI.vs.GBR)
  1. 'coefficients'
  2. 'fitted.values'
  3. 'deviance'
  4. 'method'
  5. 'unshrunk.coefficients'
  6. 'df.residual'
  7. 'design'
  8. 'offset'
  9. 'dispersion'
  10. 'prior.count'
  11. 'AveLogCPM'
  12. 'df.residual.zeros'
  13. 'df.prior'
  14. 'var.post'
  15. 'var.prior'
  16. 'samples'
  17. 'table'
  18. 'comparison'
  19. 'df.test'
  20. 'df.total'

对每个差异分析分析结果的pvalue进行校正

YRI.f.vs.m_padjust <- p.adjust(qlf_YRI.f.vs.m$table$PValue, method = "BH")
GBR.f.vs.m_padjust <- p.adjust(qlf_GBR.f.vs.m$table$PValue, method = "BH")
f.YRI.vs.GBR_padjust <- p.adjust(qlf_f.YRI.vs.GBR$table$PValue, method = "BH")
m.YRI.vs.GBR_padjust <- p.adjust(qlf_m.YRI.vs.GBR$table$PValue, method = "BH")

提取差异表达的基因

YRI.f.vs.m.sig <- qlf_YRI.f.vs.m$table[which(YRI.f.vs.m_padjust < 0.05),]
write.csv(YRI.f.vs.m.sig, file = "DE/YRI.f.vs.m_0.05_edgeR.csv")
GBR.f.vs.m.sig <- qlf_GBR.f.vs.m$table[which(qlf_GBR.f.vs.m$table$PValue < 0.05),]
write.csv(GBR.f.vs.m.sig, file = "DE/GBR.f.vs.m_0.05_edgeR.csv")
write.csv(qlf_GBR.f.vs.m$table, file = "DE/GBR.f.vs.m_edgeR.csv")

结果可视化

火山图

write.csv(counts.keep, file = "DE/counts.keep.csv")
DE <- read.csv('DE/GBR.f.vs.m_edgeR.csv',row.names = 1)

tail(DE)
logFClogCPMFPValue
ENSG00000224533 : TMLHE-AS1-0.056469327 2.052160 4.307370e-030.9486785
ENSG00000185973 : TMLHE 0.221289532 8.365665 1.286092e+000.2774115
ENSG00000168939 : SPRY3-0.002682364 4.420135 3.410488e-050.9954298
ENSG00000124333 : VAMP7-0.13738283810.762397 9.979491e-010.3361840
ENSG00000124334 : IL9R-0.488450923 6.182421 7.899469e-010.3903941
ENSG00000185203 : WASIR1-1.144990880 1.771968 5.055304e-010.4897413 
DEG <- data.frame(DE$PValue,DE$logFC)
colnames(DEG) <- c('p','logFC')
rownames(DEG) <- rownames(DE)
tail(DEG)
plogFC
ENSG00000224533 : TMLHE-AS10.9486785 -0.056469327
ENSG00000185973 : TMLHE0.2774115 0.221289532
ENSG00000168939 : SPRY30.9954298 -0.002682364
ENSG00000124333 : VAMP70.3361840 -0.137382838
ENSG00000124334 : IL9R0.3903941 -0.488450923
ENSG00000185203 : WASIR10.4897413 -1.144990880

示例数据量少，筛选出的差异基因也少，画图的效果不太好

options(repr.plot.width = 8, repr.plot.height = 6)

my_volcano <- function(DEG, t_p, t_FC = 0, t_marker){
    ## example: print(my_volcano(a[,c(5,1)], 0.01, 0.6, 0))
    library(ggplot2)
    DEG = na.omit(DEG)
    colnames(DEG) = c('p','logFC')
    DEG$gene <- unlist(lapply(strsplit(rownames(DEG),":"),function(x) x[2]))
    DEG[DEG$p < 1e-10, 'p'] = 1e-10
    
    if (t_FC == 0){
        t_FC <- with(DEG, mean(abs(logFC)) + 2 * sd(abs(logFC)))
    }
    DEG$change = as.factor(ifelse(DEG$p < t_p & abs(DEG$logFC) > t_FC,ifelse(DEG$logFC > t_FC, "UP", "DOWN"),"NOT"))
    this_title <- paste0("Cutoff for logFC is ", round(t_FC, 3),
                         "\nThe number of up gene is ",
                         nrow(DEG[DEG$change == "UP", ]),
                         "\nThe number of down gene is ",
                         nrow(DEG[DEG$change == "DOWN", ])
                        ) 
    p = ggplot(data = DEG, aes(x = logFC, y = -log10(p), color = change)) +
    geom_point() +
    scale_color_manual(values = c('blue', 'black', 'red')) +
    geom_hline(yintercept = -log10(t_p), lty = 4, lwd = 0.6, alpha = 0.8) +
    geom_vline(xintercept = c(t_FC, -t_FC), lty = 4, lwd = 0.6, alpha = 0.8) +
    theme_bw() +
    ylim(0,5) +
    xlim(-10,10) +
    theme(panel.border = element_blank(),
          panel.grid.major = element_blank(),
          panel.grid.minor = element_blank(),
         axis.line = element_line(colour = "black")) + ggtitle(this_title) +
    labs(x = "log2(fold change)", y = "-log10(p-value)") +
    theme(plot.title = element_text(hjust = 0.5))
    
    if (t_marker != 0) {
        DE <- subset(DEG, abs(logFC) > t_marker)
        p = p + geom_text(data = DE , aes(label = gene), col = "green",alpha = 0.5)
        }
    
    return(p)
}


print(my_volcano(DEG, 0.05, 1, 4.2))

Warning message:
“Removed 1 rows containing missing values (geom_point).”Warning message:
“Removed 1 rows containing missing values (geom_text).”

在这里插入图片描述

热图

suppressMessages(library(pheatmap))
suppressMessages(library(gplots))
suppressMessages(library(ComplexHeatmap))

DE_0.05 <- read.csv('DE/GBR.f.vs.m_0.05_edgeR.csv',row.names = 1)
coldata <- read.csv("geuvadis_phenodata.csv")
counts.keep <- read.csv("DE/counts.keep.csv",row.names = 1)

choose_gene=head(rownames(DE_0.05),50) 
choose_matrix=counts.keep[choose_gene,]
choose_matrix=t(scale(t(choose_matrix)))

head(choose_matrix)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
ENSG00000182378 : PLCXD1 1.9215927 0.4896378 0.9650236 -0.4910871 -1.0940155 -0.07754007 1.0742078 -1.2302541 -0.1886567 -1.2003010 -0.60123750 0.4326301
ENSG00000185291 : IL3RA 1.0684102 1.1323672 -1.1750052 -1.1405668 0.1041351 -1.07169002-0.7420654 -0.1024952 1.2750405 -1.0470912 0.78306348 0.9158973
ENSG00000196433 : ASMT 2.1365053 0.6224464 0.5215092 -0.6897379 -0.1850516 -0.89161243-0.8916124 1.5308817 -0.3869261 -0.7906752 -0.68973792-0.2859889
ENSG00000183943 : PRKX-0.2822280 -0.1640627 1.8727853 -0.8229841 -1.3016534 -1.46788583 1.3420432 -0.2101271 -0.1300151 0.2505170 -0.05190591 0.9655167
ENSG00000130021 : PUDP-0.7621578 0.2543147 2.1551890 -0.4343396 -0.9649807 -0.57820227 0.3675181 -0.8682861 -0.2739680 0.1057352 -0.67961368 1.6787912
ENSG00000101846 : STS-0.3061919 1.1965013 1.4125252 -1.2271362 -0.7609792 -0.79508827 1.5792806 -0.7041308 -0.9807931 0.1467005 -0.36872515 0.8080371 
rownames(choose_matrix) <- unlist(lapply(strsplit(rownames(choose_matrix),":"),function(x) x[2]))
head(choose_matrix)
ERR188044ERR188104ERR188234ERR188245ERR188257ERR188273ERR188337ERR188383ERR188401ERR188428ERR188454ERR204916
PLCXD1 1.9215927 0.4896378 0.9650236 -0.4910871 -1.0940155 -0.07754007 1.0742078 -1.2302541 -0.1886567 -1.2003010 -0.60123750 0.4326301
IL3RA 1.0684102 1.1323672 -1.1750052 -1.1405668 0.1041351 -1.07169002-0.7420654 -0.1024952 1.2750405 -1.0470912 0.78306348 0.9158973
ASMT 2.1365053 0.6224464 0.5215092 -0.6897379 -0.1850516 -0.89161243-0.8916124 1.5308817 -0.3869261 -0.7906752 -0.68973792-0.2859889
PRKX-0.2822280 -0.1640627 1.8727853 -0.8229841 -1.3016534 -1.46788583 1.3420432 -0.2101271 -0.1300151 0.2505170 -0.05190591 0.9655167
PUDP-0.7621578 0.2543147 2.1551890 -0.4343396 -0.9649807 -0.57820227 0.3675181 -0.8682861 -0.2739680 0.1057352 -0.67961368 1.6787912
STS-0.3061919 1.1965013 1.4125252 -1.2271362 -0.7609792 -0.79508827 1.5792806 -0.7041308 -0.9807931 0.1467005 -0.36872515 0.8080371 
options(repr.plot.width = 8, repr.plot.height = 6)

morecols <- colorRampPalette(c("green","black","red"))
col.sex <- c("purple","orange")[coldata$sex]
label_col = paste(coldata$sex ,'.', coldata$population)
annotation_col = data.frame(population = coldata$population,sex = coldata$sex)
rownames(annotation_col) = coldata$ids

ann_colors = list(population = c(GBR = "#7570B3", YRI = "#E7298A"),
                  sex = c(male = "#1B9E77", female = "#D95F02")
                 )

pheatmap(choose_matrix, color = morecols(100), labels_col = label_col, annotation_col = annotation_col,
         annotation_colors = ann_colors , fontsize_row = 6, scale="row",
         annotation_legend = T, border_color=NA,fontsize = 8, main = "genes  GBR.f.vs.m")


# pheatmap(choose_matrix, color = morecols(100), labels_col = label_col, annotation_col = annotation_col,
#          annotation_colors = ann_colors , fontsize_row = 6, scale="row",
#          annotation_legend = T, fontsize = 8, main = "genes  GBR.f.vs.m",filename = 'DEG_top50_heatmap.pdf')

在这里插入图片描述

pheatmap参数用法 ：使用pheatmap包绘制热图

my_heatmap <- function(countdata,col_label,row_lable,group){
    morecols <- colorRampPalette(c("red","black","green"))
    col.sex <- c("purple","orange")[group]
    # Plot the heatmap
    heatmap.2(countdata,labCol=col_label,srtCol=15, 
              col=rev(morecols(50)),trace="none", main="genes  GBR.f.vs.m",
              ColSideColors=col.sex,colCol=col.sex,scale="row")
}


col_label <- paste(coldata$sex ,'.', coldata$population)
group <- coldata$sex
my_heatmap(choose_matrix, col_label, row_lable, group)

在这里插入图片描述

heatmap.2详细文档

library("GetoptLong")

expr <- choose_matrix
colnames(expr) <- paste(coldata$sex ,'.', coldata$population)                     

annotation_col = data.frame(population = coldata$population,sex = coldata$sex)
rownames(annotation_col) = coldata$ids
ann_colors = list(population = c(GBR = "#7570B3", YRI = "#E7298A"),
                  sex = c(male = "#1B9E77", female = "#D95F02")
                 )
ann <- HeatmapAnnotation(annotation_col, col = ann_colors)
                       
morecols <- colorRampPalette(c("green","black","red"))                        
Heatmap(expr,heatmap_legend_param = list(title= "legend", title_position = "topcenter", 
  legend_height=unit(2,"cm"), legend_direction="vertical"), col = morecols(100), column_title = "genes  GBR.f.vs.m",
  top_annotation = ann,row_names_gp = gpar(fontsize = 6, fontface = "bold") )


#注释名称可以使用下面的R代码添加
for(an in colnames(annotation_col)) { 
seekViewport(qq("annotation_@{an}")) 
grid.text(an, unit(1, "npc") + unit(2, "mm"), 0.5, default.units = "npc", just = "left")}

在这里插入图片描述

發表評論
所有評論
還沒有人評論，想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章

BioPython-1

文件解析 針對簡單地fasta文件的解析： ls_orchid.fasta 我們知道fasta文件的一般是以>開頭，爲一個read，也就是序列的名稱，下面是序列，如這樣的： >gi|2765658|emb|Z78533.1|C

攻城猿bilibili 2020-06-24 05:01:00

HISAT2-StringTie-Ballgown有参转录组数据分析

參考文獻： Pertea M, Kim D,Pertea G M, et al. Transcript-level expression analysis of RNA-seq experimentswith HISAT, StringT

ygyxl 2020-06-23 00:18:36

使用Aspera从EBI或NCBI下载基因组数据（转）

 轉自：http://www.plob.org/2012/07/31/3013.html 做基因組數據分析，可能經常從NCBI的GEO/SRA或者EBI的ENA數據庫下載高通量的數據，動輒幾十G的數據用wget下載實在太糾結，這時就要用到

redline2005 2020-06-22 05:10:07

生物信息数据格式：fastq格式

文章目錄格式說明實例演練判斷fastq序列編碼是Phred33(Illumina1.8+) or Phred64(Illumina1.3+)fastq轉換fasta格式Linux 操作fastq獲取數據統計reads_1.fq文件

sunchengquan 2020-06-16 00:44:18

绘制heatmap

我們在分析了差異表達數據之後，經常要生成一種直觀圖－－熱圖(heatmap)。這一節就以基因芯片數據爲例，示例生成高品質的熱圖。 比如 鋼藍漸白配色的熱圖 首先還是從最簡單的heatmap開始。 > library(gg

blacklee123 2020-06-13 21:58:38

miRNA的命名规则

blacklee123 2020-02-20 17:27:28

将GEO的soft 数据转换为expression matrix

blacklee123 2020-02-20 17:27:28

vcftools用法详解

sunchengquan 2019-02-01 15:28:38

转录组分析专题

sunchengquan 2019-01-01 20:32:56

bedtools指南

sunchengquan 2018-12-16 15:04:24

生物信息数据格式：gff,gtf格式

sunchengquan 2018-12-16 15:04:13

转录组分析中的R基础知识

sunchengquan 2018-12-09 15:02:21

生信人的linux考试20题解析

sunchengquan 2018-12-07 15:10:38

序列比对

sunchengquan 2018-08-31 00:00:51

python图像库PIL画质粒

sunchengquan 2018-08-31 00:00:49