背景
在之前介绍的基因组拼接,主要是二代illumina测序的拼接,其中使用不同的软件,及每个软件对应的不同的参数,会得到不同的结果,那么怎么选取软件和参数呢,下面着重介绍这些,当然取决于你的实验数据,从实际出发。可以测试部分小数据先看看结果,拼接多了就明白了。
一、不同 kmer 大小
mkdir 1.kmer
cd 1.kmer
cp ../36.illumina/lib.list .
SOAPdenovo-31mer all -s lib.list -K 31 -o kmer31 -D 1 -d 1 -u 2 -p 12 >kmer31.log
SOAPdenovo-63mer all -s lib.list -K 63 -o kmer63 -D 1 -d 1 -u 2 -p 12 >kmer63.log
SOAPdenovo-127mer all -s lib.list -K 127 -o kmer105 -D 1 -d 1 -u 2 -p 12 >kmer105.log
ll *.scafSeq
seqkit stat *.scafSeq
mv *.scafSeq ../
rm -rf kmer*
#修改-d
SOAPdenovo-63mer all -s lib.list -K 63 -o kmer63 -D 1 -d 0 -u 2 -p 12 >kmer63_0.log
SOAPdenovo-63mer all -s lib.list -K 63 -o kmer63 -D 1 -d 1 -u 2 -p 12 >kmer63_1.log
SOAPdenovo-63mer all -s lib.list -K 63 -o kmer63 -D 1 -d 2 -u 2 -p 12 >kmer63_2.log
seqkit stat *.scafSeq
二、文库大小
小片段文库,写入文件 lib.list
#大片段文库
max_rd_len=150
[LIB]
avg_ins=500
reverse_seq=0
asm_flags=3
rank=1
pair_num_cutoff=3
q1=/cleandata/500_clean.1.fq.gz
q2=/cleandata/500_clean.2.fq.gz
[LIB]
avg_ins=2000
reverse_seq=1
asm_flags=2
rank=2
pair_num_cutoff=3
q1=/cleandata/2000_clean.1.fq.gz
q2=/cleandata/2000_clean.2.fq.gz
SOAPdenovo-63mer all -s lib.list -K 63 -o kmer1 -D 1 -d 1 -u 2 -p 12 >kmer1.log
SOAPdenovo-63mer all -s lib2.list -K 63 -o kmer2 -D 1 -d 1 -u 2 -p 12 >kmer2.log
ll *.scafSeq
seqkit stat *.scafSeq
seqkit seq -m 500 kmer1.scafSeq | seqkit stat
seqkit seq -m 500 kmer2.scafSeq | seqkit stat
换spades软件看大片段文库
echo "spades.py --isolate -o spades1 -t 12 -1 /share/home/xiehs/05.assembly/data/illumina._1.fastq.gz -2 /share/home/xiehs/05.assembly/data/illumina_2.fastq.gz -t 12 1>spades1.log 2>spades1.err" > spades.sh
echo "spades.py --isolate -o spades2 -t 12 -1 /share/home/xiehs/05.assembly/data/illumina._1.fastq.gz -2 /share/home/xiehs/05.assembly/data/illumina_2.fastq.gz -t 12 --mp1-1 /cleandata/2000_clean.1.fq.gz --mp1-2 /cleandata/2000_clean.2.fq.gz 1>spades2.log 2>spades2.err" >> spades.sh
nohup sh spades.sh &
seqkit seq -m 500 spades1/scaffolds.fasta | seqkit stat
seqkit seq -m 500 spades2/scaffolds.fasta | seqkit stat
三、数据质量
比较过滤前后拼接结果差别;
mkdir 3.filter
SOAPdenovo-63mer all -s lib.list -K 63 -o kmer63 -D 1 -d 0 -u 2 -p 12 >kmer63.log
seqkit seq -m 500 kmer63.scafSeq | seqkit stat
seqkit seq -m 500 ../1.kmer/kmer63_1.scafSeq | seqkit stat
四、数据量大小
分别抽取 10%,30%,50%,80%进行比较。
seqkit sample -p 0.1 -s 1234 /share/home/xiehs/05.assembly/data/illumina_1.fastq.gz | gzip >reads.0.1_1.fq.gz
seqkit sample -p 0.1 -s 1234 /share/home/xiehs/05.assembly/data/illumina_2.fastq.gz | gzip >reads.0.1_2.fq.gz
for i in {0.1,0.3,0.5,0.8};do
seqkit sample -p ${i} -s 1234 /share/home/xiehs/05.assembly/data/illumina_1.fastq.gz | gzip >reads.${i}_1.fq.gz
seqkit sample -p ${i} -s 1234 /share/home/xiehs/05.assembly/data/illumina_2.fastq.gz | gzip >reads.${i}_2.fq.gz;
done;
ls -1 *.fq.gz | xargs -n 2 | while read {i,j};do echo spades.py -o spades_${i} -t 12 -1 ${i} -2 ${j};done;
spades.py -o spades_reads.0.1_1.fq.gz -t 12 -1 reads.0.1_1.fq.gz -2 reads.0.1_2.fq.gz
spades.py -o spades_reads.0.3_1.fq.gz -t 12 -1 reads.0.3_1.fq.gz -2 reads.0.3_2.fq.gz
spades.py -o spades_reads.0.5_1.fq.gz -t 12 -1 reads.0.5_1.fq.gz -2 reads.0.5_2.fq.gz
spades.py -o spades_reads.0.8_1.fq.gz -t 12 -1 reads.0.8_1.fq.gz -2 reads.0.8_2.fq.gz
_1.fq.gz删掉作为目录-o
nohup sh spades.sh &
seqkit stat spades_reads*/scaffolds.fasta
五、reads 长度的影响
#不同reads长度
#利用wgsim分别模拟长度70与150bp长度reads
cp /share/home/xiehs/05.assembly/data/MGH78578.fasta .
wgsim MGH78578.fasta read.50_1.fq read.50_2.fq -1 50 -2 50
wgsim MGH78578.fasta read.300_1.fq read.300_2.fq -1 300 -2 300
spades.py -o spades50 -t 12 -1 read.50_1.fq -2 read.50_2.fq
spades.py -o spades300 -t 12 -1 read.300_1.fq -2 read.300_2.fq
seqkit stat spades50/scaffolds.fasta
seqkit stat spades300/scaffolds.fasta
六、不同错误率
#不同错误率的影响
wgsim -e 0.01 -1 50 -2 50 MGH78578.fasta read.0.01_1.fq read.0.01_2.fq
wgsim -e 0.1 -1 50 -2 50 MGH78578.fasta read.0.1_1.fq read.0.1_2.fq
#用soapdenovo拼接
SOAPdenovo-63mer all -s lib.list -K 35 -o kmer35 -D 1 -d 1 -u 2 -p 12
SOAPdenovo-63mer all -s lib.list -K 35 -o kmer35_01 -D 1 -d 1 -u 2 -p 12
seqkit stat kmer35.scafSeq
写在最后:有时间我们会努力更新的。大家互动交流可以前去论坛,地址在下面,复制去浏览器即可访问,弥补下公众号没有留言功能的缺憾。
bioinfoer.com
有些板块也可以预设为大家日常趣事的分享等,欢迎大家来提建议。