Re: Methyl-kit error with calculateDiffMeth()?

[Altuna Akalin]

In order to read BED files to GRanges, you can use the function in methylKit called read.bed(). In addition, you can get help for the functions by calling:

help(read.bed)

or 

?read.bed

Best,

Altuna

On Thursday, June 27, 2013 5:29:15 PM UTC+2, Akalin, Altuna wrote:

Hi Ian,

I am posting this reply to you and methylKit Q &A forum (methylkit_...@googlegroups.com and reachable @ https://groups.google.com/forum/#!forum/methylkit_discussion).

You need to use unite() function before calling calculateDiffMeth() function. The unite() function finds bases/regions that are shared between all samples, and only then you can use the calculateDiffMeth() function. See the example code on the web page or the vignette. 

For the second question, yes, you can look at custom regions. You need to have a GRanges object from GenomicRanges package that have the regions of interest. Then, you need to use regionCounts() function to summarize methylation information over regions. After that you need to use unite() and calculateDiffMeth()

Best,

Altuna

On Thu, Jun 27, 2013 at 11:58 AM, Ian Donaldson <Ian.Do...@manchester.ac.uk> wrote:

I have been running the latest version of MethylKit with R3, all the steps have worked until I get to calculateDiffMeth().  Then I get the errors below.

> calculateDiffMeth()
Error in (function (classes, fdef, mtable)  :
  unable to find an inherited method for function ‘calculateDiffMeth’ for signature ‘"missing"’

> meth_dup_diff = calculateDiffMeth(meth_dup_norm)
Error in (function (classes, fdef, mtable)  :
  unable to find an inherited method for function ‘calculateDiffMeth’ for signature ‘"methylRawList"’

>sessionInfo()
R version 3.0.0 (2013-04-03)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.utf8       LC_NUMERIC=C            
 [3] LC_TIME=en_GB.utf8        LC_COLLATE=en_US.utf8   
 [5] LC_MONETARY=en_GB.utf8    LC_MESSAGES=en_US.utf8  
 [7] LC_PAPER=C                LC_NAME=C               
 [9] LC_ADDRESS=C              LC_TELEPHONE=C          
[11] LC_MEASUREMENT=en_GB.utf8 LC_IDENTIFICATION=C     

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base    

other attached packages:
[1] methylKit_0.5.7

loaded via a namespace (and not attached):
[1] BiocGenerics_0.6.0   data.table_1.8.8     GenomicRanges_1.12.4
[4] IRanges_1.18.1       KernSmooth_2.23-10   parallel_3.0.0     
[7] stats4_3.0.0         tools_3.0.0        

Is there something I am obviously doing wrong?

Also I understand that differential analysis can be performed on single bases or tiled regions, but can it be done on custom regions, e.g. CpG island/shores corresponding to the Agilent SureSelectXT Human Methyl-Seq kit?

Thank you,

Ian

[Altuna Akalin]