[BioC] Question on LIMMA

[James W. MacDonald]

Hi Francisco,

Francisco Javier Burguillo wrote:
> Dear colleages,
> 
> I can not uderstand a result I have. I made t-permutations and LIMMA to
> find differentially expressed genes on the same dataset. Is it posible I 
> can have
> a gene which is overexpressed using t-permutations (t value = + something)
> and underespressed using LIMMA (Statistic= - something) ?

That's highly unlikely - the numerator (which is responsible for the 
sign of the statistic) is identical in both cases. The only differences 
are slight changes in how the denominator is computed, and what 
distribution the t-statistics are compared against.

It's more likely that you have e.g. experimental - control in one case 
and control - experimental in the other. However, without seeing  your 
code that's just a guess.

Best,

Jim


> 
> Sorry but I can not fully understand the LIMMA algorithm yet.
> 
> Many thanks,
> 
> Francisco
> 
> 
> 
> ___________________________________________
> Fco. Javier Burguillo
> Departamento de Química Física
> Facultad de Farmacia
> Universidad de Salamanca
> España
> http://simfit.usal.es
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
James W. MacDonald, M.S.
Biostatistician
Douglas Lab
University of Michigan
Department of Human Genetics
5912 Buhl
1241 E. Catherine St.
Ann Arbor MI 48109-5618
734-615-7826