生物信息学入门~利用购买的云服务器学习有参转录组数据处理（fastq到差异表达）

2024-04-30 19:40:11 浏览数 (1)

拟南芥有参转录组数据处理，基本流程是

  • 1 转录组数据过滤
  • 2 转录组数据与参考基因组比对
  • 3 基于比对结果计算count值
  • 4 基于count值进行差异表达分析

准备数据

参考基因组 拟南芥基因组 来源于 论文 The genetic and epigenetic landscape of the Arabidopsis centromeres。下载链接 https://github.com/schatzlab/Col-CEN/tree/main/v1.2

只保留染色体序列，我使用Heliex这个工具对蛋白编码基因进行了注释，获得蛋白编码基因注释的gff格式文件

转录组数据来源于论文 Rapid and Dynamic Alternative Splicing Impacts the Arabidopsis Cold Response Transcriptome

下载 时间点8 和时间点 14 的转录组数据

ERR1886187,ERR1886352,ERR1886283 timepoint8 ERR1886289,ERR1886358,ERR1886193 timepoint14

这里的数据值选取了总数据的百分之一

准备好三个输入数据

  • 1 参考基因组的fasta文件
  • 2 参考基因组的蛋白编码基因注释文件 gff格式
  • 3 转录组测序数据 fastq 格式

安装软件

这里安装软件都用conda来安装，需要用到的软件有

  • fastp
  • hisat2
  • samtools
  • stringtie
  • R语言 tidyverse包（整理数据）DEseq2 包 （差异表达分析）

数据处理

第一步：原始数据过滤

代码语言：javascript复制
fastp -i reads/ERR1886187_subset_1.fastq.gz -I reads/ERR1886187_subset_2.fastq.gz -o 01.clean.reads/ERR1886187_1.fq.gz -O 01.clean.reads/ERR1886187_2.fq.gz -j 01.fastp.output/ERR1886187.json -h 01.fastp.output/ERR1886187.html -w 1
fastp -i reads/ERR1886193_subset_1.fastq.gz -I reads/ERR1886193_subset_2.fastq.gz -o 01.clean.reads/ERR1886193_1.fq.gz -O 01.clean.reads/ERR1886193_2.fq.gz -j 01.fastp.output/ERR1886193.json -h 01.fastp.output/ERR1886193.html -w 1
fastp -i reads/ERR1886283_subset_1.fastq.gz -I reads/ERR1886283_subset_2.fastq.gz -o 01.clean.reads/ERR1886283_1.fq.gz -O 01.clean.reads/ERR1886283_2.fq.gz -j 01.fastp.output/ERR1886283.json -h 01.fastp.output/ERR1886283.html -w 1

fastp -i reads/ERR1886289_subset_1.fastq.gz -I reads/ERR1886289_subset_2.fastq.gz -o 01.clean.reads/ERR1886289_1.fq.gz -O 01.clean.reads/ERR1886289_2.fq.gz -j 01.fastp.output/ERR1886289.json -h 01.fastp.output/ERR1886289.html -w 1
fastp -i reads/ERR1886352_subset_1.fastq.gz -I reads/ERR1886352_subset_2.fastq.gz -o 01.clean.reads/ERR1886352_1.fq.gz -O 01.clean.reads/ERR1886352_2.fq.gz -j 01.fastp.output/ERR1886352.json -h 01.fastp.output/ERR1886352.html -w 1
fastp -i reads/ERR1886358_subset_1.fastq.gz -I reads/ERR1886358_subset_2.fastq.gz -o 01.clean.reads/ERR1886358_1.fq.gz -O 01.clean.reads/ERR1886358_2.fq.gz -j 01.fastp.output/ERR1886358.json -h 01.fastp.output/ERR1886358.html -w 1

第二步：对参考基因组构建索引

代码语言：javascript复制
hisat2-build fa/at.col0.chr.fna index/at

第三步：转录组数据与参考基因组进行比对

代码语言：javascript复制
hisat2 -p 1 --dta -x ref/index/at -1 01.clean.reads/ERR1886187_1.fq.gz -2 01.clean.reads/ERR1886187_2.fq.gz -S 02.sam/ERR1886187.sam
hisat2 -p 1 --dta -x ref/index/at -1 01.clean.reads/ERR1886193_1.fq.gz -2 01.clean.reads/ERR1886193_2.fq.gz -S 02.sam/ERR1886193.sam
hisat2 -p 1 --dta -x ref/index/at -1 01.clean.reads/ERR1886283_1.fq.gz -2 01.clean.reads/ERR1886283_2.fq.gz -S 02.sam/ERR1886283.sam

hisat2 -p 1 --dta -x ref/index/at -1 01.clean.reads/ERR1886289_1.fq.gz -2 01.clean.reads/ERR1886289_2.fq.gz -S 02.sam/ERR1886289.sam
hisat2 -p 1 --dta -x ref/index/at -1 01.clean.reads/ERR1886352_1.fq.gz -2 01.clean.reads/ERR1886352_2.fq.gz -S 02.sam/ERR1886352.sam
hisat2 -p 1 --dta -x ref/index/at -1 01.clean.reads/ERR1886358_1.fq.gz -2 01.clean.reads/ERR1886358_2.fq.gz -S 02.sam/ERR1886358.sam


samtools sort -@ 1 -O BAM -o 03.sorted.bam/ERR1886187.sorted.bam 02.sam/ERR1886187.sam
samtools sort -@ 1 -O BAM -o 03.sorted.bam/ERR1886193.sorted.bam 02.sam/ERR1886193.sam
samtools sort -@ 1 -O BAM -o 03.sorted.bam/ERR1886283.sorted.bam 02.sam/ERR1886283.sam

samtools sort -@ 1 -O BAM -o 03.sorted.bam/ERR1886289.sorted.bam 02.sam/ERR1886289.sam
samtools sort -@ 1 -O BAM -o 03.sorted.bam/ERR1886352.sorted.bam 02.sam/ERR1886352.sam
samtools sort -@ 1 -O BAM -o 03.sorted.bam/ERR1886358.sorted.bam 02.sam/ERR1886358.sam

第四步：计算基因表达量

代码语言：javascript复制
stringtie -p 1 -G ref/at.col0.chr.gff -e -B -o 04.stringtie.output/ERR1886187/ERR1886187.gtf -A 04.stringtie.output/ERR1886187/ERR1886187.gene_abund.tsv 03.sorted.bam/ERR1886187.sorted.bam
stringtie -p 1 -G ref/at.col0.chr.gff -e -B -o 04.stringtie.output/ERR1886193/ERR1886193.gtf -A 04.stringtie.output/ERR1886193/ERR1886193.gene_abund.tsv 03.sorted.bam/ERR1886193.sorted.bam
stringtie -p 1 -G ref/at.col0.chr.gff -e -B -o 04.stringtie.output/ERR1886283/ERR1886283.gtf -A 04.stringtie.output/ERR1886283/ERR1886283.gene_abund.tsv 03.sorted.bam/ERR1886283.sorted.bam

stringtie -p 1 -G ref/at.col0.chr.gff -e -B -o 04.stringtie.output/ERR1886289/ERR1886289.gtf -A 04.stringtie.output/ERR1886289/ERR1886289.gene_abund.tsv 03.sorted.bam/ERR1886289.sorted.bam
stringtie -p 1 -G ref/at.col0.chr.gff -e -B -o 04.stringtie.output/ERR1886352/ERR1886352.gtf -A 04.stringtie.output/ERR1886352/ERR1886352.gene_abund.tsv 03.sorted.bam/ERR1886352.sorted.bam
stringtie -p 1 -G ref/at.col0.chr.gff -e -B -o 04.stringtie.output/ERR1886358/ERR1886358.gtf -A 04.stringtie.output/ERR1886358/ERR1886358.gene_abund.tsv 03.sorted.bam/ERR1886358.sorted.bam

第五步：合并基因表达量

代码语言：javascript复制
Rscript getCounts.R

getCounts.R 的内容

代码语言：javascript复制
library(tidyverse)

myfun<-function(x){
  return(read_tsv(x) %>% 
           select(`Gene ID`,`Coverage`) %>% 
           mutate(sampleName=str_extract(x,pattern = "ERR[0-9] ")))
}
list.files("04.stringtie.output",
           pattern = "gene_abund.tsv",
           recursive = TRUE,
           full.names = TRUE) %>% 
  map(myfun) %>% 
  bind_rows() %>% 
  pivot_wider(names_from = "sampleName",values_from = "Coverage") %>% 
  write_tsv("counts.tsv")

第六步：差异表达分析

代码语言：javascript复制
library(tidyverse)
library(DESeq2)

mycounts<-read_tsv("D:/Jupyter/practice/RNAseqAT/counts.tsv") %>% 
  column_to_rownames("Gene ID") %>% 
  round()

mycounts %>% 
  head()

mycounts %>% dim()

mycounts_1<-mycounts[rowSums(mycounts) != 0,]
dim(mycounts_1)

mymeta<-read_csv("D:/Jupyter/practice/RNAseqAT/mymeta.csv")

colnames(mycounts_1) == mymeta$id

mycounts_1 %>% 
  select(mymeta$id) -> mycounts_1

colnames(mycounts_1) == mymeta$id

dds <- DESeqDataSetFromMatrix(countData=mycounts_1, 
                              colData=mymeta, 
                              design=~dex)

dds <- DESeq(dds)
res <- results(dds)
res


res %>% 
  as.data.frame() %>% 
  mutate(group = case_when(
    log2FoldChange >= 1 & padj <= 0.05 ~ "UP",
    log2FoldChange <= -1 & padj <= 0.05 ~ "DOWN",
    TRUE ~ "NOT_CHANGE"
  )) %>% 
  pull(group) %>% table()


res %>% 
  as.data.frame() %>% head()

res %>% 
  as.data.frame() %>% 
  mutate(group = case_when(
    log2FoldChange >= 1 & padj <= 0.05 ~ "UP",
    log2FoldChange <= -1 & padj <= 0.05 ~ "DOWN",
    TRUE ~ "NOT_CHANGE"
  )) %>% 
  ggplot(aes(x=log2FoldChange,y=-log10(padj))) 
  geom_point(aes(fill=group),size=8,shape=21,
             color="gray",
             alpha=0.6) 
  scale_fill_manual(values = c("UP"="red",
                                "DOWN"="blue",
                                "NOT_CHANGE"="gray")) 
  theme_bw(base_size = 20) 
  theme(panel.grid = element_blank(),
        legend.position = "top")
数据 云服务器 数据处理 output 入门

0 人点赞