[R-sig-ME] Gene expression lme

[Federico Calboli]

On 26 Apr 2012, at 16:30, Angelina Mukherjee wrote:
> 
> I have 2 patients, with around 20,000 probes each. Each patient has
> observations of gene expression (response) taken from some regions of their
> tissue. Then the next level in the hierarchy of the data is the sub region
> - we have data from several subregions for each region of each patient. So
> I have 3 levels of nesting : subregion within region within patient.
> 
> Here, A, B, C and D correspond to subregions of patient 1 and together they
> comprise Region 1 for the first patient. GHIJ comprise the next and so on.
> In such a situation, is the following the correct syntax for the lmer
> command?
> ( I am considering patient, region and subregion to be random factors. )
> 
> *fit <- lmer(expression ~ 1  + probe + (1|patient)+ (1|patient:region) +
> (1|patient: region: subregion) )*

If each subregion is uniquely identified, so that subregion A can only be associated with patient X, and subregion Z can only be found in patient Y (and similarly, the same is done for regions), I though that

lmer(epression ~ probe + (1|patient) + (1|region) + (1|subregion)) 

would work.  Unique identifiers for the same region in different patients might sound odd at first, but in my experience it works and simplifies the specification of the nesting, which then happens automatically



> 
> The problem is that this encounters the same problem as lme. The whole data
> is of the dimension 20,000 x 22, where 20,000 is the number of probes and
> 22 is the number of samples/arrays.
> 
> Any advice would be very helpful!
> 
> Cheers,
> Angelina
> 
>> 
>> 
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

--
Federico C. F. Calboli
Neuroepidemiology and Ageing Research
Imperial College, St. Mary's Campus
Norfolk Place, London W2 1PG

Tel +44 (0)20 75941602   Fax +44 (0)20 75943193

f.calboli [.a.t] imperial.ac.uk
f.calboli [.a.t] gmail.com