[BioC] Using GenomeGraphs effectively for generating publication quality pictures

[steffen at stat.Berkeley.EDU]

Hi Daren,

> 1) How to include the cytoband information e.g. 4q to the ideogram ? How
> to label the gene in the cytoband with "IL8" ?

Currently you can not include cytoband information on the ideogram.   How
would you want this to look like?

> 2) What colour is actually used by gene and transcript ? I have created a
> legend with "orange" and "lightblue", but the colour doesn't look the
> same.

You can set your own colors for these by the DisplayPars function when
creating genes and transcripts, e.g.:

gene <- new("Gene", id = "TP53", type="hgnc_symbol", biomart = mart,
dp=DisplayPars(color="green"))


By default the colors are orange for genes and cornflowerblue for
transcripts.

> 3) How to make the legend a single line ? Currently, the words are below
> the boxes

This is hard-coded at this point.

> 4) I have affy U133A, exon 1.0ST and agilent 44K probes to map to the
> gene. How can I do that ?

I think all these platforms are in Ensembl so you could use biomaRt to
link them to the Ensembl Gene id. Here's an example:

library(biomaRt)
mart = useMart("ensembl", dataset="hsapiens_gene_ensembl")
map = getBM(c("ensembl_gene_id", "affy_hg_u133a"), filters =
"affy_hg_u133a", values=c("209718_at","221633_at"), mart=mart)

>map
  ensembl_gene_id affy_hg_u133a
1 ENSG00000025770     209718_at
2 ENSG00000025770     221633_at


> 5) How to space the numbers away from the genomeAxis so that it looks less
> clustered ?

This is hard-coded at this time as well.

Hope this helps.
Steffen

>
> Hi,
>
> I am using GenomeGraphs to generate gene structures. My codes are below:
>
> A few questions that I have
> 1) How to include the cytoband information e.g. 4q to the ideogram ? How
> to label the gene in the cytoband with "IL8" ?
> 2) What colour is actually used by gene and transcript ? I have created a
> legend with "orange" and "lightblue", but the colour doesn't look the
> same.
> 3) How to make the legend a single line ? Currently, the words are below
> the boxes
> 4) I have affy U133A, exon 1.0ST and agilent 44K probes to map to the
> gene. How can I do that ?
> 5) How to space the numbers away from the genomeAxis so that it looks less
> clustered ?
>
>
> library("biomaRt")
> library("GenomeGraphs")
>
> mart <- useMart("ensembl", dataset="hsapiens_gene_ensembl")
> drawGS <- function(gene_symbol) {
>   gs <- getBM(attributes=c("hgnc_symbol", "ensembl_gene_id",
> "chromosome_name", "start_position", "end_position", "band"),
> filter=c("hgnc_symbol"), values=gene_symbol, mart=mart)
>
>   title <- new("Title", title = paste(gs[1], ":", gs[2]), dp =
> DisplayPars(color = "darkred"))
>   ideog <- new("Ideogram", chromosome = as.character(gs[3]))
>   genomeAxis <- new("GenomeAxis", add53 = TRUE, add35 = TRUE)
>   gene <- new("Gene", id = as.character(gs[2]), biomart = mart)
>   transcript <- new("Transcript", id = as.character(gs[2]), biomart =
> mart)
>
>   legend <- new("Legend", legend = c("gene", "transcript"), dp =
> DisplayPars(color = c("orange", "lightblue")))
>
>   pdf(paste(gene_symbol, ".pdf", sep=""))
>   gdPlot(list(title, ideog, genomeAxis, gene, transcript, legend), minBase
> = as.integer(gs[4]), maxBase = as.integer(gs[5]))
>   dev.off()
> }
>
> drawGS("IL8")
> _________________________________________________________________
> [[elided Hotmail spam]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
> http://news.gmane.org/gmane.science.biology.informatics.conductor
>