[BioC] iranges

[Valerie Obenchain]

Hi,

Please 'reply all' when responding so communication stays on the list.

If you are working with stranded ranges you should use the GRanges 
container. IRanges is not strand-aware and does not have a strand 
argument. You can see the function signature on the man page by typing

?IRanges

> Usage:
>
>      ## IRanges constructor:
>      IRanges(start=NULL, end=NULL, width=NULL, names=NULL)
>

Load a Transcript Db object and extract transcripts by gene:
library(TxDb.Dmelanogaster.UCSC.dm3.ensGene)
tx <- transcriptsBy(TxDb.Dmelanogaster.UCSC.dm3.ensGene, "gene")

Select a gene with transcripts on the negative strand:
gene <- tx[[3]]
>> gene
> GRanges with 4 ranges and 2 metadata columns:
>       seqnames               ranges strand |     tx_id     tx_name
>          <Rle>            <IRanges>  <Rle> | <integer> <character>
>   [1]    chr3R [12632936, 12655767]      - |     21863 FBtr0306337
>   [2]    chr3R [12633349, 12653845]      - |     21864 FBtr0083388
>   [3]    chr3R [12633349, 12655300]      - |     21865 FBtr0083387
>   [4]    chr3R [12633349, 12655474]      - |     21866 FBtr0300485

GRanges can be manipulated with resize(), trim(), shift(), flank(), 
narrow() and several other methods. To see them type (with the quotes)

?`intra-range-methods`

and select the page for GRanges. It sounds like resize() is what you're 
looking for.

resize(gene, width = 10)
>> resize(gene, width = 10)
> GRanges with 4 ranges and 2 metadata columns:
>       seqnames               ranges strand |     tx_id     tx_name
>          <Rle>            <IRanges>  <Rle> | <integer> <character>
>   [1]    chr3R [12655758, 12655767]      - |     21863 FBtr0306337
>   [2]    chr3R [12653836, 12653845]      - |     21864 FBtr0083388
>   [3]    chr3R [12655291, 12655300]      - |     21865 FBtr0083387
>   [4]    chr3R [12655465, 12655474]      - |     21866 FBtr0300485

If you have sequence data instead of range data, the XStringSet family 
is more appropriate. For examples of manipulating sequences see Section 
E on the XStringSet man page. The functions you want are narrow() or 
subseq().

library(Biostrings)
?XStringSet


Valerie


On 08/08/2014 08:38 AM, carol white wrote:
> I have the problem when i want to take the width from the end of a
> sequence on a reverse strand.
> if I take the nucleotide seq of a gene that is on the reverse strand on
> the ncbi web site and extract for ex 10 or 20 bp from the end, i don't
> get the same as I do with iranges. As I have already given the strand as
> the parameter to the iranges function, I assume that it has already
> reverse-complemented by iranges. I don't have this problem with the
> genes that are on the forward strand nor when I take the sub sequence
> from the beginning of the sequence.
>
> Regards,
> On Friday, August 8, 2014 5:28 PM, Valerie Obenchain
> <vobencha at fhcrc.org> wrote:
>
>
> Did you provide 'start', 'end' and 'width' and get a confusing answer?
> If yes, please show your example.
>
> Thanks.
> Valerie
>
>
>
> On 08/08/2014 08:23 AM, Valerie Obenchain wrote:
>  > Hi Carol,
>  >
>  > The 'end' is the end of the range. When you specify ranges with 'end'
>  > and 'width' the range will always end at the 'end' value.
>  >
>  >  > IRanges(end = 10, width = c(5, 10))
>  > IRanges of length 2
>  >      start end width
>  > [1]    6  10    5
>  > [2]    1  10    10
>  >
>  >
>  > Similar reasoning for 'start' and 'width':
>  >
>  >  > IRanges(start = 10, width = c(5, 10))
>  > IRanges of length 2
>  >      start end width
>  > [1]    10  14    5
>  > [2]    10  19    10
>  >
>  >
>  > Valerie
>  >
>  >
>  >
>  > On 08/08/2014 01:29 AM, carol white wrote:
>  >> Hi,
>  >> How does width with start and end in IRanges work? I thought that if I
>  >> use the end with a width, then the sequence from the end with the
>  >> length of width is taken. However, in my case when I use width for ex
>  >> 20 and 10, the corresponding sequences with the length 20 and 10 are
>  >> not the same from the end but from the beginning. Did I misunderstood
>  >> some thing?
>  >>
>  >> Regards,
>  >>
>  >> Carol
>  >>    [[alternative HTML version deleted]]
>  >>
>  >> _______________________________________________
>  >> Bioconductor mailing list
>  >> Bioconductor at r-project.org <mailto:Bioconductor at r-project.org>
>  >> https://stat.ethz.ch/mailman/listinfo/bioconductor
>  >> Search the archives:
>  >> http://news.gmane.org/gmane.science.biology.informatics.conductor
>  >>
>  >
>  >
>
>
> --
> Valerie Obenchain
> Program in Computational Biology
> Fred Hutchinson Cancer Research Center
> 1100 Fairview Ave. N, Seattle, WA 98109
>
> Email: vobencha at fhcrc.org <mailto:vobencha at fhcrc.org>
> Phone: (206) 667-3158
>
>