Error with annotateWithGeneParts command

[Saima Rehman]

Hi!

I am trying to annotate the differentially methylation regions in the samples using the genomation package. I downloaded a UCSC bed file (txt format). I used the following settings in the UCSC browser: 

Using this file I made a gene.obj using the following code:

gene.obj=readTranscriptFeatures("zebrafish.txt")

The gene.obj looks like this:

myDiff25p object looks like this:

After this, I put the annotateWithGeneParts command as follows:

> annotateWithGeneParts(as(myDiff25p_all,"GRanges"),gene.obj)

I get the following error

Error in h(simpleError(msg, call)) : 

  error in evaluating the argument 'x' in selecting a method for function 'nearest': error in evaluating the argument 'x' in selecting a method for function 'ranges': subscript contains NAs

In addition: Warning messages:

1: In .Seqinfo.mergexy(x, y) :

  The 2 combined objects have no sequence levels in common. (Use

  suppressWarnings() to suppress this warning.)

2: In .Seqinfo.mergexy(x, y) :

  The 2 combined objects have no sequence levels in common. (Use

  suppressWarnings() to suppress this warning.)

3: In .Seqinfo.mergexy(x, y) :

  The 2 combined objects have no sequence levels in common. (Use

  suppressWarnings() to suppress this warning.)

4: In .Seqinfo.mergexy(x, y) :

  The 2 combined objects have no sequence levels in common. (Use

  suppressWarnings() to suppress this warning.)

5: In .Seqinfo.mergexy(x, y) :

  The 2 combined objects have no sequence levels in common. (Use

  suppressWarnings() to suppress this warning.)

6: In .Seqinfo.mergexy(x, y) :

  The 2 combined objects have no sequence levels in common. (Use

  suppressWarnings() to suppress this warning.)

I think the error has something to do with "different chromosome naming conventions used for the annotation file and methylation call file". But I am not sure how to get around this error.

Any help in how to solve this problem would be much appreciated.

Regards,

Saima

[Altuna Akalin]

Your chromosome names are different from UCSC. That could be the problem 

--
You received this message because you are subscribed to the Google Groups "methylkit_discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to methylkit_discus...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/methylkit_discussion/654e865b-59ec-43c7-a592-9d7a5cfcc710n%40googlegroups.com.

--

Sent from mobile, excuse the brevity

[Kristen Krolick]

Hello, 

Saima, what did you end up doing to fix this issue? I am having the same issue where my chromosome names are different.

Thank you,

Kristen

[Sangeeta Kumari]

I have the same issue

[Alexander Blume]

You have to find an annotation that matches the reference assembly that your reads were aligned against.
If your assembly is not available on UCSC, see this previous post.

Best,
Alex