需要准备的文件是序列信息bed格式文件或者GRanges格式文件
比如有一个bed文件：
$ cat test.bed
chr1 1000010 1000020
chr2 1000021 1000030
chr3 1000031 1000040
我习惯保存为GR格式：
> test = read.table("test.bed")
> gr = toGR(test)
> gr
GRanges object with 3 ranges and 0 metadata columns:
seqnames ranges strand
[1] chr1 1000010-1000020 *
[2] chr2 1000021-1000030 *
[3] chr3 1000031-1000040 *
-------
seqinfo: 3 sequences from an unspecified genome; no seqlengths
toGR是我自己编写的函数，这里就不放出来了。此外还编写了另外两个函数，分别为：
-
getSeq，用于得到碱基序列
getSeq <- function(POSs){
suppressPackageStartupMessages(library(BSgenome.Hsapiens.UCSC.hg19))
seqs = c()
for(i in 1:length(POSs)){
POS = POSs[i]
chr = sapply(strsplit(POS, "[-:]"), function(z) z[1])
strat = as.numeric(sapply(strsplit(POS, "[-:]"), function(z) z[2]))
end = as.numeric(sapply(strsplit(POS, "[-:]"), function(z) z[3]))
seqs = c(seqs, as.character(BSgenome.Hsapiens.UCSC.hg19[[chr]][strat:end]))
}
return(unlist(seqs))
}
-
printSeq，用于打印碱基序列
printSeq <- function(Seq, binwidth = 50){
N_line = nchar(Seq) / binwidth
if(N_line > as.integer(N_line)){
N = as.integer(N_line) + 1
} else{
N = as.integer(N_line)
}
bases = strsplit(Seq, "")[[1]]
for(i in 0:(N-1)){
xseq = paste0(bases[(1 + binwidth*i):(binwidth + i*binwidth)], collapse = "")
if(i != (N-1)){
xseq = paste0(bases[(1 + binwidth*i):(binwidth + i*binwidth)], collapse = "")
} else{
xseq = paste0(bases[(1 + binwidth*i):length(bases)], collapse = "")
}
cat(xseq, "\n")
}
}
默认在fasta中每行显示50个碱基
然后使用前面的GR数据生成fasta文件：
> seqs = getSeq(as.character(gr))
> seqs
[1] "CTCACCCAGGA" "AAATTGAAGA" "AGAAGGAAAC"
> ID = as.character(seqnames(gr))
> ID
[1] "chr1" "chr2" "chr3"
sink("test.fa")
for(i in 1:length(seqs)){
cat(paste0(">", ID[i]), "\n")
printSeq(seqs[i])
}
system("sed 's/ //g' test.fa > tmp.fa")
system("rm test.fa")
system("mv tmp.fa test.fa")
sink()
这样就生成了fasta文件了：
$ cat test.fa
>chr1
CTCACCCAGGA
>chr2
AAATTGAAGA
>chr3
AGAAGGAAAC
最后还需要创建索引：
system("samtools faidx test.fa")
这样就完成了~
好家伙，samtools有提取序列的功能，还能指定每行的碱基数，位点格式为：chr:start-end，例如
$ samtools faidx hg19.fa chr1:2066354-2066580
>chr1:2066354-2066580
AGGCAGGGGACAGACGGACCCGGCCTGCGTTGGCCTGGGGTGACTTCACGGCTCCACTGT
CAGCAAGCGGCCGTCCCGTGGTGGATCCTGTCCGCCCTGCGAGGACACCTGGCTCCATCC
ACACCTGGGCCTCTGTCTCCAGCCGCCGAGGCCGTGACACCATGAGGATCATGTGAGGAG
GGGCAGAGAGAGGCCTCCGGGAGGCCGTCATTCCAGCCCTGCCTTCC
还能将想要提取的区域写入文件中，并传递给-r参数。每个区域单独一行
区域的格式刚好是Granges对象非常容易生成的格式，所以可以写个GR转fasta的小函数：
grToFasta <- function(gr, file){
hg19.fa = "hg19.fa"
pos = as.character(gr)
write.table(pos, "tmp.pos", sep = "\n", quote = FALSE, col.names = FALSE, row.names = FALSE)
system(paste("samtools faidx -n 50", hg19.fa, "-r tmp.pos >", file))
system(paste("rm tmp.pos & samtools faidx", file))
}
我直接好家伙，还是调包香~