Target Prediction Module - Empty result files in miRGate_results

[Matti Meikäläinen]

Hi,

I am doing target prediction analysis of DE miRNAs as decribed in the manual in section 4.1.7.3

Everything has worked fine until Target Prediction Module. I have several comparisons and the EdgeR result files contain significant miRNAs.

I have used these parameters

[TargetPrediction]
noiseq_cutoff=0.8
edger_cutoff=0.05

But the module is returning only empty files to miRGate_results with only the header:

#miRNA  miRNA FC    miRNA FDR   Ensembl Gene    Gene Symbol Ensembl Transcript  Gene FC Gene FDR    Method  Target Site Score   Energy

No errors were reported, and when I put the list of DE miRNAs here

http://mirgate.bioinfo.cnio.es/miRGate/

it generates a table.

Could this have something to do with the university proxy policies again?

Thank you!

[Matti Meikäläinen]

Actually, now I can't even access the site.

Proxy Error

The proxy server received an invalid response from an upstream server. The proxy server could not handle the request GET /.

Reason: Error reading from remote server

Apache/2.2.14 (Ubuntu) Server at mirgate.bioinfo.cnio.es Port 80

[Matti Meikäläinen]

But now I can. Do you know what is going on?

[Matti Meikäläinen]

And one question: from the manual I got the impression that one can also do Target Prediction using just miRNAs, or alternatively with both miRNAs and mRNAs, or is the module not working because I need them both.

[Eduardo Andres Leon]

Dear Matti, did you receive my emails ?

Eduardo

[Matti Meikäläinen]

Hi Eduardo,

No, I didn't. Perhaps some settings prohibit it.

[Eduardo Andres Leon]

Fine :( . I saw today the same with another user.

I attach you a screenshot of my mails

[Matti Meikäläinen]

I noticed something fishy in the summary and EdgeR part: I can explain it with the example summary file:

This is how it should look:

But in my summary, there are some empty cells in the Comparison column like this:

I checked my contrastfile and it looks OK... And so does sampleinfo.

And the other columns (File, etc.) have the same information like that.

And then, after EdgeR nothing is in the summary file, but the Target Prediction Module still keeps running and producing those empty files. And "finished successfully".

I have 31 comparisons myself and the EdgeR directory contains 31 result xls files, but the summary file has 62 rows like that in the EdgeR part... Everything else is OK:

[Eduardo Andres Leon]

Are you using the 1.6.1 miARma version ?

I suppose, according whit this last information, that you are not running the examples ?

[Matti Meikäläinen]

Yes, I am not running the examples. And after re-running this again, the Edge results look OK in the summary.xls, but still miRGate failed. And I am using v1.6.1 of miARma. So if I just use the web page and give the lists there and get the results from there. http://mirgate.bioinfo.cnio.es/miRGate/

And btw, can I do functional analysis easily with just these miRGate results if they are not directly from the pipeline, but from the miRGate site?

[Eduardo Andres Leon]

But, did you have miRGate results for the examples ??? I guess that your proxy/network settings doesn't allow to use the miRGate API. The result file format form the web page is complete "different" from the one in the API (that is needed for miARma). Besides, the web page is intended for a few genes/miRNAs while the API has no limitations

[Matti Meikäläinen]

I was just showing the example summary file previously. But I have now tried miARma-Seq on my own desktop and miRGate works. So the proxy policies are causing this, as I suspected from the beginning already. Thank you for all your help.

[Eduardo Andres Leon]

Glad to know that your error is not related with miARma

You`r welcome

[Matti Meikäläinen]

Just one more question: can I do functional analysis with the miRGate results? The miRNA pipeline here http://miarmaseq.cbbio.es/usage ends to TargetPrediction, and Functional Analysis [FAnalysis] is only for the mRNA example and before TargerPrediction, but in the workflow Functional Analysis follows Target Prediction?

[Eduardo Andres Leon]

No. The functional Analysis is only available for analysed mRNAs, because the expression is corrected by the length bias. So for predicted mRNAs we can not offer this feature. But I recommend you to use miRPath 3 or miRSystem

[Matti Meikäläinen]

At least it didn't work for me, because I got an error:

[Tue May 16 09:17:40 2017] Starting a Functional Analysis.
Problem while running this R command:
        source("/lib/CbBio/RNASeq/R-Scripts/F_Analysis.R")
        setwd("xxxxx/miARmaSeq_RNA_1_130517/Functional_Analysis_results")
        resultsfiles<-NA
        resultsfiles<-F_Analysis(projectdir="xxxxx/miARmaSeq_RNA_1_130517/Functional_Analysis_results",up="xxxxx/miARmaSeq_RNA_1_130517/Functional_Analysis_results/.up_entities_edgeR.txt",down="xxxxx/miARmaSeq_RNA_1_130517/Functional_Analysis_results/.down_entities_edgeR.txt",universe="xxxxx/miARmaSeq_RNA_1_130517/Functional_Analysis_results/.universe_entities_edgeR.txt",organism="human",method="edgeR",seq_id="edger_cutoff",label="miARmaSeq_RNA_1_130517_cut_bw1_Comp_CuO1")


Error:
file(filename, "r", encoding = encoding) : 
  cannot open the connection
Calls: source -> file
In addition: Warning message:
In file(filename, "r", encoding = encoding) :
  cannot open file '/lib/CbBio/RNASeq/R-Scripts/F_Analysis.R': No such file or directory
Execution halted

[Matti Meikäläinen]

ok thank you!