[R] wgcna

[Fabrice Tourre]

Peter,

Thank you. In fact, I am also very interesting to WGCNA.

On Sat, Jun 16, 2012 at 3:29 AM, Peter Langfelder
<peter.langfelder at gmail.com> wrote:
> On Fri, Jun 15, 2012 at 8:04 AM, Ingezz <irholtman at gmail.com> wrote:
>> Dear Peter,
>>
>> I am trying to apply the WGCNA meta-analysis for two (or more) microarray
>> datasets-tutorial to my own data.
>>
>>> mp=modulePreservation(multiExpr,multiColor,referenceNetworks=1,verbose=3,networkType="signed",
>>> nPermutations=30,maxGoldModuleSize=100,maxModuleSize=400)
>>
>> However, the error I am getting is:
>>
>> Error in .checkExpr(multiData, verbose, indent) :
>>  The submitted 'multiExpr' data contain genes or samples
>>   with zero variance or excessive counts of missing entries.
>>   Please use the function goodSamplesGenes on each set to filter out the
>> problematic
>>   genes and samples before running modulePreservation.
>>
>> Seems pretty clear, but applying goodSamplesGenes function results in no
>> exclusions for both arrays.
>>
>>>  GM2 <- goodSamplesGenes(M2, minFraction = 1/2, minNSamples =8, minNGenes
>>> =20)
>>
>> What could be the issue here? Should I increase the stringency? What
>> stringency should be used with goodsamplegenes for modulepreservation to
>> work?
>
> Hi Inge,
>
> first, I suggest that in the future you contact me directly, since
> others on this list probably have no idea what we're talking about.
>
> It is difficult to diagnose the problem without the data. I suggest
> you specify the verbose argument to both functions with a high value
> (e.g., 5) which should produce more informative output.
>
> I assume you are aware of the fact that the function goodSamplesGenes
> returns a list whose component allOK determines whether any genes or
> samples should be removed. Other components in the list specify which
> genes and/or samples should be removed. The actual removal needs to be
> done by you (I haven't written a function to do that yet but may do so
> in the future to make it more convenient). In this sense the error
> message is a bit misleading and I will change that as well.
>
> HTH,
>
> Peter
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.