[BioC] (no subject)

Marta Agudo martabar at um.es
Tue Jul 12 16:17:31 CEST 2005

Hi James
Thank you

Your first point  solves a big problem I had: where is the zero signal in an
array?. You gave me the answer: nobody knows so we ought to compare  

And the second,   why do I think it is useful  to know that? Imagine that we
have the system ready to do this kind of analysis:

 will say in liver, in healthy conditions, no caspase 3 is expressed, or
yes, that in normal situation, in spite of what has been described, caspase
3 is present. This tells you apoptosis  is ready to be triggered even when
the liver is in a good shape.

May be in brain carbohydrate metabolism RNAs are highly represented whereas
lipid ones are low represented, thus  giving  you information about what a
tissue/cell etc needs and uses for its welfare

With these naive expression data,  you can tell a gene is expressed de novo
vs naive, not just upregulated but newly expressed.

It would be like  having the tissue/cell/organ gene expression standards. 


Marta Agudo PhD
Departamento de Oftalmología
Facultad de Medicina
Campus Espinardo
30100 Murcia- Spain
Phone:+34 968363996

-----Mensaje original-----
De: James W. MacDonald [mailto:jmacdon at med.umich.edu] 
Enviado el: martes, 12 de julio de 2005 15:53
Para: Marta Agudo
CC: Bioconductor at stat.math.ethz.ch
Asunto: Re: [BioC] (no subject)

Hi Marta,

I don't think this idea makes much sense scientifically for at least two 
reasons, and probably more.

1.) How exactly will you distinguish genes that are expressed from those 
that are not expressed? Note that in canonical microarray analyses 
nobody is claiming that a certain gene is expressed or not, only that it 
is expressed at a different level in one sample vs another.

2.) If you could somehow accurately determine which genes are being 
expressed, of what use is that information? When you are comparing two 
samples, you know phenotypically what the differences are, so you can 
attribute (rightly or wrongly) the differences in expression to that 
phenotypic difference. If you are just looking at e.g., normal liver and 
you find 5000 genes that are expressed, how do you attribute those genes 
to any phenotype or process (other than to note the trivial result that 
the liver appears to express these 5000 genes)?



Marta Agudo wrote:
> Hi there
> I´ve been thinking about gene expression in just one condition without
> comparing to anything else. 
> I explain better: I have data from an  affy array experiment using naive
> tissue RNA, and I want to know which genes, out of the 30000 present in
> chip, are being expressed in this tissue.  
> I would like to know  is this analysis is  possible, i mean not just
> statistically but also if scientifically has any sense,
> And if it is I would need some help
> a) is it possible to use bioconductor and GCRMA analysis ? then,  anyone
> knows a script or could guide me?
> b) how many replicas  do we need? 
> c) which is the cut off point?
> Basically which are the pros and the cons of this kind of analysis?
> thank you very much!
> marta
> Marta Agudo PhD
> Departamento de Oftalmología
> Facultad de Medicina
> Campus Espinardo
> 30100 Murcia- Spain
> Phone:+34 968363996
> 	[[alternative HTML version deleted]]
> ------------------------------------------------------------------------
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor

James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109

More information about the Bioconductor mailing list