Last point I forget is about keeping in your selected genes those that are with
a fold change > 2, a nice p value, but a signal that remains in the "noise" or
background for instance is you have in condition 1 a median/average signal at 20
then in condition 2 a median/average signal at 80; then you have a 4 fold
increase, you can get a significant adjusted p value for this, but what is the
significance of this gene if the background/noise is let say around 100 ? I m
not sure this kind of gene is useful in the model.
Regards




Philippe G


-----E-mail d'origine-----
De : Lilian <lijoli@...>
A : microarray@yahoogroups.com
Envoyé le : Mercredi, 1 Juillet 2009 0:26
Sujet : [microarray-group] Microarray analysis advise needed!!! Thanks


































Hello all!



I have ran an microarray experiment analysis comparing monocytes from blood to
macrophages from endometrium from pregnant cows. I used the bovine agilent chip
4x44K, and my experimental design is showed below. I tried to used the
R-Bioconductor to analyze my data but there is no annotation package for the
bovine agilent array, also I tried the LimmaGUI, but I could not use the
LimmaGui because of my technical replications were read as a biological
replications. So, I used the JMP genomics to analyze my data. My raw data was
previous normalized within array using=2
0the LOWESS normalization and in the JMP I performed the quantile normalization
to normalize the data between arrays. I used the proc ANOVA to do the stats and
find the differential regulated genes.

 PROC ANOVA

Model: Animal, Replicate, Tissue

Fixed Effects: Tissue, Replicate, Replicate*Tissue

Random effects: Animal, Animal*Tissue, Animal*Replicate, Animal*Tissue*Replicate

 

I used the FDR correction fixed at 0.01 the genes were considered
differentially expressed (DE)  had p-value<0.05 and at least 2-fold increased
ou decreased expression.

 

I have got around DE 1,200 genes and now I am working on the ontology analysis.
I have a problem because the agilent ID is not read in most of the softwares for
gene ontology and in general I get only 300 genes annotated out of over 600
genes that I analyze using the DAVID. Do you have any comments on that? Is that
normal not get the ontology analysis for more than a half genes?

 

I am really new on microarray analysis and I am working by myself, it has been a
big challenge for me to do it, so I really appreciate any comments that you
might have.

 

Thanks in advance

 

Lilian Oliveira

 

Lilian Oliveira DVM Ms

Ph.D. Candidate

Animal Molecular and Cell Biology

University of Florida

Phone/Fax: (352) 392-5590/392-5595



Replicate   Tissue      Channel

1                1        
      Blood       Cy3

1                1              Endo       
CY5

2                1              Blood      
Cy3

2                1              Endo       
CY5

3                1              Blood      
Cy3

3                1              Endo       
CY5

4                1              Blood      
Cy3

4                1              Endo       
CY5

1                2              Endo       
Cy3

1           0    2              Blood       Cy5

2                2              Endo       
Cy3

2                2              Blood      
Cy5

3                2              Blood      
Cy3

3                2              Endo       
Cy5

4                2              Endo       
Cy3

4                2              Blood      Cy5



[Non-text portions of this message have been removed]




















=2
0
















________________________________________________________________________
Reçevez AOL Mail sur votre téléphone. Vos e-mails accessibles à tout moment!
Créez un e-mail gratuit aujourd’hui.



[Non-text portions of this message have been removed]