Thanks Henrik. I had definitely been confused about tags. I see they are a method for adding attributes to output objects (datasets) - neat.
I had wanted to generalize a question about the QuantileNormalization typesToUpdate argument, > qn <- QuantileNormalization(csBC, typesToUpdate="pm") and thought bringing up tags was the way to do it. So much for that - sorry for the confusion. I'm still not clear on the typesToUpdate argument - I will submit another question on that. _Taku On Jan 28, 1:52 pm, Henrik Bengtsson <henrik.bengts...@aroma- project.org> wrote: > Hi. > > I've changed the subject of this thread, because the term "probe tag" > is misleading/ambiguous in the context of aroma. > > On Fri, Jan 28, 2011 at 1:22 PM, Taku Tokuyasu <tok...@gmail.com> wrote: > > Is there a way to get the list of probe category tags used by > > QuantileNormalization, and what they denote? E.g. I have been using > > qn <- QuantileNormalization(csBC, typesToUpdate="pm") > > I think you're misunderstanding how the term "tag" is defined. Data > sets and a data files (items in data sets) have names and tags. > Models/methods process data sets and they generate new data sets often > with new tags appended, cf. > > http://aroma-project.org/definitions/ > > The above method will take the data set 'csBC' and do quantile > normalization on it based on and updating only the PM probes (argument > typesToUpdate="pm"). The generated output data set will have a tag > appended to the existing ones. Yes, that tag will be "pm" because you > used setting typesToUpdate="pm", but it does not mean that argument > 'typesToUpdate' specifies a tag. > > > I've also seen e.g. "mainOnly" in the docs. > > The "mainOnly" *tag* is from an example I posted in one of my replied. > I think it was something like: > > qn <- QuantileNormalization(csR, typesToUpdate="pm", tags="*,mainOnly") > > That is just telling the method that the output data set should have > tag "mainOnly" appended to the default ones ("*"). The following > would be equivalent and maybe more clear: > > qn <- QuantileNormalization(csR, typesToUpdate="pm", tags=c("*", "mainOnly")) > > What is happening here is that the use adds a "custom" tag. What it > is does not change the function what so ever. The only tag that is > interpreted by aroma is "*", which is expanded to be the default set > of tags. > > > I guess they are coming > > from the cdf somehow, or constructed by the user - any hints on how to > > get/manipulate these would appreciated. > > Yes, what is defined as a perfect-match (PM) or a mismatch (MM) probe > is given by the CDF file. > > > Also, getTags() on various > > aroma objects did not coming up with anything (yet). > > As said above, data sets and data files have tags and those tags can > be queried using getTags() on such objects. > > > > > I see that in the general case, units to quantile normalize can be > > specified by unit name or index, as in > > http://groups.google.com/group/aroma-affymetrix/browse_thread/thread/... > > Yes, you can either specify which probes (we use the term "cell" in > the aroma framework) you want to estimate the QN model on and/or which > you want to update. You can specify the *type* of probes, e.g. PM, > MM, all. See for instancehttp://aroma-project.org/node/141. > Alternatively, you can specify the indices of the probes, which the > thread 'Normalizing only a subset of the units (Was: Re: > [aroma.affymetrix] Re: Question for custom CDF of ST-Array)' you are > referring to are doing. That thread also shows how to identify those > cell indices when you only know the unit names. > > > > > OK, I'll step back to a yet more newbie question: > > RmaBackgroundCorrection() adjusts only PM probes by default. What is > > the best way to get documentation on the function, so I can see what > > PM means here, and what other options are available? > > help("RmaBackgroundCorrection") - you may also have to follow "..." > arguments to see what is passed internally (here to > BackgroundCorrection). From the help you also see that internally > bg.adjust() of affy package is used for the actually background > correction. > > > Will the code > > break if Affy no longer tags their probes as PM? etc. > > That will never happen. Basically all probes on a chip are > perfect-match probes unless they are mismatch probes (and the latter > Affymetrix has been discarding in recent generations of chip types). > > Hope this helps > > /Henrik > > > > > _Taku > > > -- > > 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 with websitehttp://www.aroma-project.org/. > > To post to this group, send email to aroma-affymetrix@googlegroups.com > > To unsubscribe and other options, go tohttp://www.aroma-project.org/forum/ > > -- 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 with website http://www.aroma-project.org/. To post to this group, send email to aroma-affymetrix@googlegroups.com To unsubscribe and other options, go to http://www.aroma-project.org/forum/