Thanks Henrik. That was actually quite helpful. The median may indeed be a quick & easy solution.
Cheers, Steve Henrik Bengtsson wrote: > A quick comment: > > yes, the fitPLM() of affyPLM is rather flexible, but we don't have a > transparent way for supporting all its features. It is not too hard > to provide your own aroma.affymetrix wrapper though; people have done > it. However, if you omit the probe affinities you will end fitting: > > log2(y_ik) = log2(theta_i) + eps_ik > > A non-robust estimate is then > > log2(theta_i)* = mean_k(log2(y_ik)) > > A robust estimate is: > > log2(theta_i)* = median_k(log2(y_ik)) > > which is equivalent to: > > theta_i* = median_k(y_ik) > > and this estimator is provided by: > > plm <- AvgPlm(cs); > > Note that it, just like RmaPlm(), will fit the model for each unit > group separately. There is currently no implementation for doing > "mergeGroups=TRUE" of this one. > > Does this help > > Henrik > > PS. Yes, I've been quiet - still rather busy and I hope to get back to > answer all questions soon; the code development is still going forward > though. > > On Sun, Apr 5, 2009 at 8:18 PM, Steve Pederson > <stephen.peder...@student.adelaide.edu.au> wrote: >> Hi, >> >> I'm trying to fit the probelevel model for RMA processed Exon Array >> data, and would like to omit the probe affinities term from the model. >> >> If I was using the fitPLM command directly I would be able to specify >> something like >> Pset <- fitPLM(data, model = PM ~ -1 + samples) >> >> Is there a way of specifying this using the aroma.affymetrix framework? >> The only argument I can see that vaguely corresponds to this would be >> "probeModel" in the ProbeLevelModel class, but this obviously isn't able >> to accept models of this form. >> >> The only other way that I can think of would be to fit the model for >> each array separately, which would then omit this term from the model by >> default. >> >> Do any of the more seasoned hands have any other suggestions? >> >> Thanks in advance, >> >> Steve >> > > > > --~--~---------~--~----~------------~-------~--~----~ When reporting problems on aroma.affymetrix, make sure 1) to run the latest version of the package, 2) to report the output of sessionInfo() and traceback(), and 3) to post a complete code example. You received this message because you are subscribed to the Google Groups "aroma.affymetrix" group. To post to this group, send email to aroma-affymetrix@googlegroups.com To unsubscribe from this group, send email to aroma-affymetrix-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/aroma-affymetrix?hl=en -~----------~----~----~----~------~----~------~--~---
begin:vcard fn:Steve Pederson n:Pederson;Steve org:University Of Adelaide;Discipline Of Paediatrics adr:72 King William Rd;;Women's & Children's Hospital;North Adelaide;SA;5006;Australia email;internet:stephen.peder...@student.adelaide.edu.au title:Ph.D Candidate tel;work:+618 8161 6729 tel;cell:+614 1333 9618 version:2.1 end:vcard
