2. Nanopore三代测序fastq基因组比对
因为预处理会过滤掉90%以上的reads,所以分别用原始数据和过滤后的数据进行基因组比对,看是否有差异。
一、先用未经过预处理的fastq文件进行比对(Test)
- 查看文件里有多少读段
cat BC07.fastq|grep "^@"|wc -l
818523
cat BC08.fastq|grep "^@"|wc -l
502760
- 建立比对数据库
(1)准备基因序列和基因注释文件
-
从Ensemble数据库下载小鼠基因组序列
GRCm39 (GCA_000001635.9) -
从Gencode网站下载基因注释GTF文件
GRCm39 Comprehensive gene annotation(GHR) -
解压基因组序列和基因注释
gunzip Mus_musculus.GRCm39.dna_sm.primary_assembly.fa.gz
gunzip gencode.vM36.annotation.gtf.gz
(2)Hisat2比对RNA-seq数据
师兄说Hisat2并不是一个很好用的比对工具,对于三代测序的长reads,minimap2才是更好的比对工具。
- 安装Hisat2软件
conda install -c bioconda hisat2
- 使用Hisat2构建索引
hisat2-build -p 8 Mus_musculus.GRCm39.dna_sm.primary_assembly.fa GRCm39_index
GRCm39_index → 生成的索引前缀(将创建多个索引文件)
- Hisat2比对单端reads
hisat2 -p 8 -x ../../../genome/mouse_genome/GRCm39_index -U BC07.fastq -S BC07.sam
802807 reads; of these:
802807 (100.00%) were unpaired; of these:
802224 (99.93%) aligned 0 times
462 (0.06%) aligned exactly 1 time
121 (0.02%) aligned >1 times
0.07% overall alignment rate
hisat2 -p 8 -x ../../../genome/mouse_genome/GRCm39_index -U BC08.fastq -S BC08.sam
479463 reads; of these:
479463 (100.00%) were unpaired; of these:
479102 (99.92%) aligned 0 times
348 (0.07%) aligned exactly 1 time
13 (0.00%) aligned >1 times
0.08% overall alignment rate
#比对效率极低
- 比对结果分析(Samtools)
(1)samtools生成bam二进制文件,并建立索引
samtools view -@ 8 -bS BC07.sam | samtools sort -@ 8 -o BC07_sorted.bam
samtools index BC07_sorted.bam
samtools view -@ 8 -bS BC08.sam | samtools sort -@ 8 -o BC08_sorted.bam
samtools index BC08_sorted.bam
(2)通过samtools输出比对的统计信息
samtools flagstat BC07_sorted.bam
802965 + 0 in total (QC-passed reads + QC-failed reads)
802807 + 0 primary
158 + 0 secondary
0 + 0 supplementary
0 + 0 duplicates
0 + 0 primary duplicates
741 + 0 mapped (0.09% : N/A)
583 + 0 primary mapped (0.07% : N/A)
0 + 0 paired in sequencing
0 + 0 read1
0 + 0 read2
0 + 0 properly paired (N/A : N/A)
0 + 0 with itself and mate mapped
0 + 0 singletons (N/A : N/A)
0 + 0 with mate mapped to a different chr
0 + 0 with mate mapped to a different chr (mapQ>=5)
samtools flagstat BC08_sorted.bam
479481 + 0 in total (QC-passed reads + QC-failed reads)
479463 + 0 primary
18 + 0 secondary
0 + 0 supplementary
0 + 0 duplicates
0 + 0 primary duplicates
379 + 0 mapped (0.08% : N/A)
361 + 0 primary mapped (0.08% : N/A)
0 + 0 paired in sequencing
0 + 0 read1
0 + 0 read2
0 + 0 properly paired (N/A : N/A)
0 + 0 with itself and mate mapped
0 + 0 singletons (N/A : N/A)
0 + 0 with mate mapped to a different chr
0 + 0 with mate mapped to a different chr (mapQ>=5)
#为什么比对效率这么低呢?
-
IGV可视化
(1)打开IGV软件,选择Mouse基因组IGV(2)导入:File>Load from file...>.bam文件(.bam.bai要和.bam文件在同一目录下)
- 因为比对上的序列数量太少了,用samtools查看比对序列在染色体上的位置
samtools idxstats BC07_sorted.bam
1 195154279 30 0
10 130530862 15 0
11 121973369 29 0
12 120092757 210 0
13 120883175 11 0
14 125139656 14 0
15 104073951 20 0
16 98008968 23 0
17 95294699 122 0
18 90720763 9 0
19 61420004 6 0
2 181755017 26 0
3 159745316 12 0
4 156860686 25 0
5 151758149 16 0
6 149588044 17 0
7 144995196 26 0
8 130127694 25 0
9 124359700 34 0
MT 16299 45 0
X 169476592 8 0
Y 91455967 0 0
samtools view BC07_sorted.bam | awk '$3=="12" {print $3, $4}' | sort | uniq -c | sort -nr | head -n 10
77 12 69206069
13 12 69408101
9 12 69408102
7 12 69408093
4 12 69408136
4 12 69206071
3 12 69408139
3 12 69408134
3 12 69408111
3 12 69408108
-
查看第12号染色体上比对的位置,把位置复制到IGV中即可查看
所以IGV可视化可以看到比对上的都是短片段,不是长片段,这可能就和Hisat2的工具的不适合有关系。
二、用经过预处理的fastq文件进行比对(Test)
预处理fastq来源于“1. Nanopore三代测序fastq数据预处理”中的*output.fastq文件
- 查看预处理文件的reads数
cat BC07_output.fastq |grep "^@"|wc -l
105907
cat BC08_output.fastq |grep "^@"|wc -l
33509
- 进一步去掉可能的rRNA污染(可省略)
- 安装SortMeRNA
conda install -c bioconda sortmerna
mkdir -p ~/sortmerna_db && cd ~/sortmerna_db
wget -c https://github.com/biocore/sortmerna/raw/master/data/rRNA_databases/silva-bac-16s-id90.fasta
wget -c https://github.com/biocore/sortmerna/raw/master/data/rRNA_databases/silva-bac-23s-id98.fasta
wget -c https://github.com/biocore/sortmerna/raw/master/data/rRNA_databases/silva-euk-18s-id95.fasta
wget -c https://github.com/biocore/sortmerna/raw/master/data/rRNA_databases/silva-euk-28s-id98.fasta
wget -c https://github.com/biocore/sortmerna/raw/master/data/rRNA_databases/rfam-5s-database-id98.fasta
wget -c https://github.com/biocore/sortmerna/raw/master/data/rRNA_databases/rfam-5.8s-database-id98.fasta
-
如果下载的特别慢,则手动下载rRNA参考基因组,保存在新建的sortmerna_db文件夹中
silva-bac-16s-id90.fasta
silva-bac-23s-id98.fasta
silva-euk-18s-id95.fasta
silva-euk-28s-id98.fasta
rfam-5s-database-id98.fasta
rfam-5.8s-database-id98.fasta -
sortmerna做rRNA基因组比对
sortmerna --ref silva-bac-16s-id90.fasta /
--ref silva-bac-23s-id98.fasta /
--ref silva-euk-18s-id95.fasta /
--ref silva-euk-28s-id98.fasta /
--ref rfam-5s-database-id98.fasta /
--ref rfam-5.8s-database-id98.fasta /
--reads ../nanopore-data/FMR1/barcode-241126/BC08_output.fastq /
--aligned ../nanopore-data/FMR1/barcode-241126/BC08_output_rRNA /
--other ../nanopore-data/FMR1/barcode-241126/BC08_output_non_rRNA /
--fastx
sortmerna --ref silva-bac-16s-id90.fasta /
--ref silva-bac-23s-id98.fasta /
--ref silva-euk-18s-id95.fasta /
--ref silva-euk-28s-id98.fasta /
--ref rfam-5s-database-id98.fasta /
--ref rfam-5.8s-database-id98.fasta /
--reads ../nanopore-data/FMR1/barcode-241126/BC07_output.fastq /
--aligned ../nanopore-data/FMR1/barcode-241126/BC07_output_rRNA /
--other ../nanopore-data/FMR1/barcode-241126/BC07_output_non_rRNA /
--fastx
- 再次检查处理后的reads数
cat BC07_output_non_rRNA.fq |grep "^@"|wc -l
68184
cat BC08_output_non_rRNA.fq |grep "^@"|wc -l
21547
BC08去掉了60%的rRNA,BC07去掉了20%的rRNA,但是BC07是input样品,可能BC08就是富集了rRNA的mRNA,所以考虑不用去除rRNA基因组的fastq文件,直接使用*output.fastq文件。
- Hisat2比对RNA-seq数据
hisat2 -p 8 -x ../../../genome/mouse_genome/GRCm39_index -U BC08_output.fastq -S BC08_output.sam
32834 reads; of these:
32834 (100.00%) were unpaired; of these:
32682 (99.54%) aligned 0 times
144 (0.44%) aligned exactly 1 time
8 (0.02%) aligned >1 times
0.46% overall alignment rate
hisat2 -p 8 -x ../../../genome/mouse_genome/GRCm39_index -U BC07_output.fastq -S BC07_output.sam
105179 reads; of these:
105179 (100.00%) were unpaired; of these:
104630 (99.48%) aligned 0 times
350 (0.33%) aligned exactly 1 time
199 (0.19%) aligned >1 times
0.52% overall alignment rate
- 预处理和不预处理基因组比对的reads比较:

- samtools生成bam二进制文件,并建立索引
samtools view -@ 8 -bS BC07_output.sam | samtools sort -@ 8 -o BC07_output_sorted.bam
samtools index BC07_output_sorted.bam
samtools view -@ 8 -bS BC08_output.sam | samtools sort -@ 8 -o BC08_output_sorted.bam
samtools index BC08_output_sorted.bam
- IGV可视化(同前)

主要问题:
- 经过质量控制后能比对上的reads比例上升了,但是总数下降了
- 不知道是不是RIP建库的问题,基因组比对的方法能比对上的reads数太太太太少了
- 按同样的流程把BC07和BC08的重复数据再跑一遍pipeline。
共有 0 条评论