2020-05-08 复现GSE137675 RNA-Seq结果(1)：上游分析

创建项目

cd /data/project
mkdir -p GSE137675 && cd GSE137675
workdir=/data/project/GSE137675 

1. 下载原始FastQ数据

cd ${workdir}
mkdir -p fq
cd fq

1.1 使用循环创建下载地址文件

for i in {78..89}
do
echo "fasp.sra.ebi.ac.uk:/vol1/fastq/SRR101/0${i}/SRR101397${i}/SRR101397${i}_1.fastq.gz"; 
echo "fasp.sra.ebi.ac.uk:/vol1/fastq/SRR101/0${i}/SRR101397${i}/SRR101397${i}_2.fastq.gz"; 
done > fileID

1.2 编写批量下载脚本aspera.sh

cat fileID | while read id 
do
~/.aspera/connect/bin/ascp -QT -l 50m -P33001  \
-i ~/.aspera/connect/etc/asperaweb_id_dsa.openssh   \
era-fasp@${id}  .
done

1.3 运行脚本aspera.sh批量下载FastQ文件

nohup bash aspera.sh 1>../aspera.log 2>&1 &

2. 对原始测序FastQ文件进行质量评估

conda activate rna
cd ${workdir}
nohup fastqc -t 6 -o ./qc/ ./fq/*.gz 1>fastqc.log 2>&1 &
multiqc -o ./qc/ ./qc/*

3. 对FastQ文件进行过滤

conda activate rna
cd ${workdir}
mkdir clean && cd clean
ln -s /data/project/GSE137675/fq/*.gz ./

3.1 编写批量运行脚本fastp.sh

for i in $(ls *_1.fastq.gz)
do
    i=${i%_1.fastq.gz}
    fastp -i ${i}_1.fastq.gz -o ${i}_1.fastp.fq.gz \
        -I ${i}_2.fastq.gz -O ${i}_2.fastp.fq.gz \
        -l 18 -q 20 --compression=6 \
        -R ${i} -h ${i}.fastp.html -j ${i}.fastp.json \
        1>${i}.fastp.log 2>&1
done

3.2 运行fastp.sh脚本批量修剪过滤

nohup bash fastp.sh 1>../trim.log 2>&1 &

3.3 对fastp修剪过滤后的结果质量评估

conda activate rna
cd ${workdir}
mkdir clean_qc
nohup fastqc -t 6 -o ./clean_qc/ ./clean/*.fastp.fq.gz  1>clean.log 2>&1 &
cd clean_qc
multiqc -o ./ *.zip

4. 比对到参考基因组

conda activate rna
cd ${workdir}
makdir -p mapping && cd mapping
ln -s /data/project/GSE137675/clean/*.fastp.fq.gz ./

4.1 编写批量比对脚本hisat2.sh

index=/data/reference/index/hisat2/hg38/hg38/genome
ls *_1.fastp.fq.gz | while read id
do
id=${id/_1.fastp.fq.gz/}; 
hisat2 -p 6 -x ${index} -1 ${id}_1.fastp.fq.gz \
-2 ${id}_2.fastp.fq.gz 2>${id}.hisat2.log | \
samtools sort -@ 6 -o ${id}.hisat2.sort.bam -
done

4.2 运行hisat2.sh脚本批量比对

nohup bash hisat2.sh 1>../mapping.log 2>&1 &

4.3 文献中存在一个过滤条件比对质量>=20，编写过滤脚本filter.sh

ls *.sort.bam | while read id
do
id=${id%.hisat2.sort.bam}
samtools view -@ 6 -q 20 -o ${id}.sort.filter.bam ${id}.hisat2.sort.bam 
done

4.4 运行filter.sh脚本批量过滤

nohup bash filter.sh 1>../filter.log 2>&1 &

4.5 对bam文件进行统计

ls *.filter.bam | while read id
do 
samtools flagstat -@ 6 ${id} > ./flagstat/${id/bam/flagstat}
done

5. featureCounts计数

conda activate rna
cd ${workdir}
mkdir -p counts && cd counts/
ln -s /data/project/GSE137675/mapping/*.filter.bam ./
ls *.bam | while read id; do mv ${id} ${id%.sort.filter.bam}; done

gtf=/data/reference/gtf/hg38/gencode.v33.annotation.gtf.gz
nohup featureCounts -T 6 -p \
-t exon -g gene_id \
-a $gtf -o all.id.txt * \
1>../counts.log 2>&1 &
cut -f1,7- all.id.txt > counts.txt
cat counts.txt | less -S

image.png