序列比对之Biostrings包

基本概念

Biostrings包很重要的3个功能是进行Pairwise sequence alignment 和Multiple sequence alignment及 Pattern finding in a sequence

序列比对一般有2个过程：

1）构建计分矩阵公式（the scoring matrix formulation)

2）比对(alignment itself)

global alignment methods （全局比对）：align every  residue in the sequences ，例如Needleman-Wunsch algorithm.

local alignment technique（局部比对）： align regions of high similarity in the sequences，例如Smith-Waterman algorithm

安装

if("Biostrings" %in% rownames(installed.packages()) == FALSE) {source("http://bioconductor.org/biocLite.R");biocLite("Biostrings")}
suppressMessages(library(Biostrings))
ls('package:Biostrings')

----------------Pairwise sequence alignment---------------

步骤：首先构建罚分规则，然后按照规则进行比对。用pairwiseAlignment（）函数

举例1：核酸序列

(myScoringMat <- nucleotideSubstitutionMatrix(match = 1, mismatch = -1, baseOnly = TRUE))#构建罚分规则
gapOpen <- 2             #gap分为2
gapExtend <- 1           #延伸gap分为1

sequence1 <- "GAATTCGGCTA"  #序列1
sequence2 <- "GATTACCTA"    #序列2
myAlignment <- pairwiseAlignment(sequence1, sequence2,
                                 substitutionMatrix = myScoringMat, gapOpening = gapOpen,
                                 gapExtension = gapExtend, type="global", scoreOnly = FALSE)  #进行比对
myAlignment

举例2：对蛋白序列进行比对

蛋白比对会更复杂，因此模型更多，

data(package="Biostrings")  #查看所有数据集
data(BLOSUM62)               #这里选择BLOSUM62数据
subMat <- "BLOSUM62"         #赋值
gapOpen <- 2
gapExtend <- 1
sequence1 <- "PAWHEAE"
sequence2 <- "HEAGAWGHE"
myAlignProt <- pairwiseAlignment(sequence1, sequence2,
                                 substitutionMatrix = subMat, gapOpening = gapOpen, gapExtension =
                                   gapExtend, type="global", scoreOnly = FALSE)  #全局比对
myAlignProt2 <- pairwiseAlignment(sequence1, sequence2,
                                 substitutionMatrix = subMat, gapOpening = gapOpen, gapExtension =
                                   gapExtend, type="local",scoreOnly = FALSE)  ##局部比对

可以看到局部比对返回的是，高度相似的序列部分.

3)可视化，对于序列可以用最经典的对角线来可视化（以人和黑猩猩的hemoglobin beta为例）

library(seqinr) # 为了读取fasta序列
myseq <- read.fasta(file = "F:/R/Bioconductor/biostrings/prtein_example_seq.fas")
dotPlot(myseq[[1]], myseq[[2]], col=c("white", "red"), xlab="Human", ylab="Chimpanzee")

##########Multiple sequence alignment############

一般多序列比对可以用于进化分析

install.packages("muscle")  #需要安装该包,因为该包在我的版本上没法安装，所以这里就不讲了
library(muscle)

######Phylogenetic analysis and tree plotting########

这里先不做分析

#########blast格式的解析######

install.packages("RFLPtools",dependencies=TRUE)
library(RFLPtools)
data(BLASTdata)    #先查看数据集了解一下相关数据格式情况
head(BLASTdata)
colnames(BLASTdata)
DIR <- system.file("extdata", package = "RFLPtools") #用自带数据集
MyFile <- file.path(DIR, "BLASTexample.txt")
MyBLAST <- read.blast(file = MyFile)
mySimMat <- simMatrix(MyBLAST)   #可以根据blast结果用来生成相似性矩阵，太厉害了

#########Pattern finding in a sequence######

library(Biostrings)
mynucleotide <- DNAString("aacataatgcagtagaacccatgagccc")
matchPattern(DNAString("ATG"), mynucleotide)  #示例1
matchPattern("TAA", mynucleotide)       #示例2
##以下函数可以用来寻找orf（需要修改）
myCodonFinder <- function(sequence){
  startCodon = DNAString("ATG") # 指定起始密码子
  stopCodons = list("TAA", "TAG", "TGA") # 指定终止密码子
  codonPosition = list() #initialize the output to be returned as a list
  codonPosition$Start = matchPattern(startCodon, sequence) # search start codons
  x=list()
  for(i in 1:3){ # iterate over all stop codons
    x[[i]]= matchPattern(DNAString(stopCodons[[i]]), sequence)
    codonPosition$Stop=x
  }
  return(codonPosition) # returns results
}
StartStops <- myCodonFinder(mynucleotide)