Hii, i'm trying to analyse Two colors Agilent Microarrays with Robin. Seems to be very easy to introduce data. But when i choose one option to perform the quality control an error message apears on the monitor.

dyld: Library not loaded: /usr/local/lib/libgfortran.2.dylib
Referenced from: /Users/evalima/Desktop/Robin.app/Contents/Resources/Java/R/lib/R/bin/exec/R
Reason: image not found

I cannot understant why this is happening.

I'm using MAC OS X version 10.6.8
and the Robin Version 1.1.2

Thank you in advance

Eva Oliveira

Hi Eva,

I'm just checking, was this solved by email?
I just found this thread by accident as unanswered!

Axel

Hiii,

No one contact me about my question.

Coulb you please help me with this error!!!

Hi Eva Lima Oliveira,

you are using a quite old version of Robin - please consider
installing the latest version from our web site - this should already
solve you problem.

bests,
Marc

Thanks, now is working perfectly.

I'm processiding with the analysis and the problem that i'm facing now is the design of the experiment. Because the tutorial that i have is an old version that not correspond to the one that i have.

I'm using the version 1.3.2 for MAc, and the manual that i have is from march 2010.

Can you send me the newst manual??

Thank you in advance.

Eva

Hi Eva,

the manual available for download right now dates from May 2012 - However,
the microarray part is the same as in older versions (with some minor changes
and additions) since the latest addition was the RNA-Seq analysis workflow.
So it might not really help you with your experiment design problem. If you give
us more details of your problem, we might be able to help.

cheers,
Marc

Hiii,

i'll try to explain my problem. But first i introduce the design of array's maybe this could be usefull.

I have 3 biological replicates of one assay. We did two color agilent array 44K.

With cy3 we labbel the control and with cy5 labbel the treated. The hybridization was performed with the same control sample.

So to perform the meta groups i do like this:

cBAP-BAP1
cBAP-BAP2
cBAP-BAP3

an then the error appears!

[1] "C"
Read /Users/evalima/Desktop/Robin_BAP/input/array09.txt.robin
Read /Users/evalima/Desktop/Robin_BAP/input/array10.txt.robin
Read /Users/evalima/Desktop/Robin_BAP/input/array20.txt.robin
[1] no duplicate spots detected.
[1] inserting unique ID according to chip layout.
Found unique target names:
BAP1 BAP2 BAP3 cBAP
Warning message:
Partial NA coefficients for 1701 probe(s)
Error in ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim) :
No residual degrees of freedom in linear model fits
Calls: eBayes -> ebayes
Execution halted


And i don't now whats wrong!!!

Thank you in advance!

Eva Oliveira

Dear Eva Oliveira,

if i understand you correctly, you have used three arrays in two color mode to
measure differential expression between two conditions in three biological replicates
(please correct me if I'm wrong). Looking at the error message i guess that
you defined the experiment incorrectly - if cBAP is one condition and BAP the other,
you should not define the three replicate BAP samples as "BAP1", "BAP2" and "BAP3"
but simply as BAP and set up the targets table to reflect which color channel
on which chip contains which sample. So (provided I understood you correctly)
your targets table should look sth. like this:


/Users/evalima/Desktop/Robin_BAP/input/array09.txt.robin cBAP BAP
/Users/evalima/Desktop/Robin_BAP/input/array10.txt.robin cBAP BAP
/Users/evalima/Desktop/Robin_BAP/input/array20.txt.robin cBAP BAP


That would be a simple design of three replicate measurements of two conditions
without dye swap. If this reflects your actual experiment's design please try to
rerun RobiNA with the targets table set up like above. The error message you got implies
that there is no replication in your experiment (hence the "no residual degrees of
freedom" message). But this is just because the targets table has to be defined in
a different way.

I hope this solves your problem - with best greetings,
Marc

Thanks for your message.

You understand exactly the design of my experiment. Is a simple 3 biological replicate experiment. Only two conditions.
But now i'm facing a new error message.

After the quality analysis summary page, I click in the next bottom and an error message apears:

[1] "C"
Error in file(file, "r") : cannot open the connection
Calls: read.maimages -> switch -> readGenericHeader -> file
In addition: Warning message:
In file(file, "r") : cannot open file 'NA': No such file or directory
Execution halted

You have any suggestion to do??

Thanks in advance!

Eva Oliveira

Hi Eva Oliveira,

the error message by itself is unfortunately quite generic and does not allow me to
see where / why exactly the problem occurs - could you maybe send me the complete
analysis script that was generated by RobiNA so that i can have a detailed look?

The script file can be found in the "source" subfolder of your project directory and will
be called sth like "EXPERIMENT_NAME_main_analysis.R".

You can send the file via email (lohse_at_mpimp-golm.mpg.de) rather than posting it
here.

cheers,
Marc

Hi Eva,

the targets table looks perfectly alright, but after taking a closer look at the analysis
script, it occurred to me that the chip layout that you specified cannot be correct.

You state in your email that you are working with AGILENT 44k arrays but the layout
that was specified defines only 1 spot. When importing the data, please make sure to
select the proper chip layout in the "chip layout" section of the import dialog. You need to
specify the number of grid columns and rows and the number of spots per row and column
within each block. There should be a AGILENT 44k preset that you can choose from the
dropdown box - if there's none, you would have to specify the layout yourself and then
save it for future use (so that you don't have to type it in every time). As far as i know,
Agilent chips have no grid design so you'd end up with 1 grid row, 1 grid column, 532 rows
per grid and 85 columns per grid (532 * 85 = 45220 = 44k spots).

I am surprised that the quality checking step did not fail - nevertheless, since some of the
checks only work correctly if the layout was correctly specified, you'll have to discard
any quality check results that you already obtained and re-do quality analysis with the
layout given above.

I hope this solves your problem.

cheers,
Marc