CompoSeq {GeneR} | R Documentation |
Gives composition in mono, di or trinucleotides of fragments of a sequence
Only A, T, G, C are counted, letter as U, X, S, M, N, are counted as 'X'. Flag "case" specify weather we take into acount letters in lower, upper ar all case.
compoSeq(seqno=0, from=1, to=0, strand=getStrand(), wsize=1, p=TRUE,case="all",step=wsize) strCompoSeq(s, wsize=1, p=TRUE,case="all",step=wsize)
seqno |
Integer/scalar, Sequence number (buffer number) |
s |
String/scalar, A sequence as character string |
wsize |
Integer/scalar, size of k-uples. |
from,to |
Integer/scalar, Absolute addresses of the beginning and the end of the fragment, (1 means the first nucleotide and 0 conventionally the last one; from must not be larger than to and both vectors must be the same size) |
strand |
Integer/vector, Strand (forward: 0, reverse: 1) |
p |
Logical/scalar, If p=TRUE, results are in pourcentages, else, results are in counts. |
case |
String/scalar, Case of the letters taken into account ("all", "upper", "lower") |
step |
Integer/scalar, Step increment of positions when counting words. |
A matrix with nucleotide composition for all regions (NxM where N = 5*wordSize and M = length(from))
If error: NULL
.Use setStrand by default
s<-"CGTACGTAGTAGCTAGCTAGCTAGCTAGCTGATCGATGCTAGCTGATCGATGCT" placeString(s) x<-c(1,8,15,50) y<-c(5,12,19,54) compoSeq(from =x, to = y,wsize=2) compoSeq(from =x, to = y,wsize=2,strand=1) compoSeq(from =x, to = y,wsize=2,step=2,p=FALSE)