[BioC] subset in XPS

[Zhibin Lu]

Dear Christian,

The new functions are working great!

I really appreciate your help,

Zhibin

> Date: Sat, 5 Jul 2008 21:49:24 +0200
> From: cstrato at aon.at
> To: zhbluweb at hotmail.com
> CC: bioconductor at stat.math.ethz.ch
> Subject: Re: [BioC] subset in XPS
>
> Dear Zhibin
>
> I have just uploaded a new version xps_1.1.3 to BioC devel:
> http://bioconductor.org/packages/2.3/bioc/html/xps.html
> which hopefully satisfies your requests:
>
> 1. to allow CEL-names starting with a number, "check.names=FALSE"
> is added to read.table() in all methods.
>
> 2. functions root.expr() and root.call() are added to allow
> loading ExprTreeSet and CallTreeSet from a root file
> (just like load SchemeTreeSet and DataTreeSet).
>
> Thus, you can now subset directly from a root file:
>> sub.rma <- root.expr(scheme.test3, "Test3RMA.root", "mdp",
> c("TestA2", "TestB1"))
>
> Please let me know if the new version solves your problems.
>
> Best regards
> Christian
>
>
> Zhibin Lu wrote:
>> Dear Christian,
>>
>> I tested the new package with my linux box, so I compiled ROOT from source with the file 'root_v5.18.00.source.tar.gz'. Since I was using Linux, I think xps was also compiled from source. I followed your direction, it seems that the function 'exprs' gave the extra X. Here is what I got:
>>
>>
>>> treenames<-treeNames(Data.rma)
>>> treenames
>>>
>> [[1]]
>> [1] "10017_12.mdp"
>>
>> [[2]]
>> [1] "100M_11.mdp"
>>
>> [[3]]
>> [1] "11017_4.mdp"
>>
>> [[4]]
>> [1] "110M_3.mdp"
>>
>> [[5]]
>> [1] "11117_6.mdp"
>>
>> [[6]]
>> [1] "111M_5.mdp"
>>
>> [[7]]
>> [1] "9617_2.mdp"
>>
>> [[8]]
>> [1] "96M_1.mdp"
>>
>> [[9]]
>> [1] "9717_8.mdp"
>>
>> [[10]]
>> [1] "97M_7.mdp"
>>
>> [[11]]
>> [1] "9817_10.mdp"
>>
>> [[12]]
>> [1] "98M_9.mdp"
>>
>>
>>> treenames=getTreeNames("rma_all.root")
>>> treenames
>>>
>> [1] "10017_12.rbg" "10017_12.int" "100M_11.rbg" "100M_11.int" "11017_4.rbg"
>> [6] "11017_4.int" "110M_3.rbg" "110M_3.int" "11117_6.rbg" "11117_6.int"
>> [11] "111M_5.rbg" "111M_5.int" "9617_2.rbg" "9617_2.int" "96M_1.rbg"
>> [16] "96M_1.int" "9717_8.rbg" "9717_8.int" "97M_7.rbg" "97M_7.int"
>> [21] "9817_10.rbg" "9817_10.int" "98M_9.rbg" "98M_9.int" "10017_12.cqu"
>> [26] "100M_11.cqu" "11017_4.cqu" "110M_3.cqu" "11117_6.cqu" "111M_5.cqu"
>> [31] "9617_2.cqu" "96M_1.cqu" "9717_8.cqu" "97M_7.cqu" "9817_10.cqu"
>> [36] "98M_9.cqu" "10017_12.mdp" "100M_11.mdp" "11017_4.mdp" "110M_3.mdp"
>> [41] "11117_6.mdp" "111M_5.mdp" "9617_2.mdp" "96M_1.mdp" "9717_8.mdp"
>> [46] "97M_7.mdp" "9817_10.mdp" "98M_9.mdp"
>>
>>
>>> treenames=getTreeNames("rma_all.root", "mdp")
>>> treenames
>>>
>> [1] "10017_12.mdp" "100M_11.mdp" "11017_4.mdp" "110M_3.mdp" "11117_6.mdp"
>> [6] "111M_5.mdp" "9617_2.mdp" "96M_1.mdp" "9717_8.mdp" "97M_7.mdp"
>> [11] "9817_10.mdp" "98M_9.mdp"
>>
>>
>>> value<-exprs(Data.rma)
>>> treenames<-colnames(value)
>>> treenames
>>>
>> [1] "UNIT_ID" "UnitName" "X10017_12.mdp_LEVEL"
>> [4] "X100M_11.mdp_LEVEL" "X11017_4.mdp_LEVEL" "X110M_3.mdp_LEVEL"
>> [7] "X11117_6.mdp_LEVEL" "X111M_5.mdp_LEVEL" "X9617_2.mdp_LEVEL"
>> [10] "X96M_1.mdp_LEVEL" "X9717_8.mdp_LEVEL" "X97M_7.mdp_LEVEL"
>> [13] "X9817_10.mdp_LEVEL" "X98M_9.mdp_LEVEL"
>>
>>
>>> treenames <- treenames[c(4,6)]
>>> sub.rma<-Data.rma
>>> exprs(sub.rma, treenames) <- value
>>> str(sub.rma)
>>>
>>
>> ..@ treenames:List of 2
>> .. ..$ : chr "X100M_11.mdp"
>> .. ..$ : chr "X110M_3.mdp"
>>
>>
>> Regards,
>>
>> Zhibin
>>
>>
>>> Date: Thu, 3 Jul 2008 21:58:38 +0200
>>> From: cstrato at aon.at
>>> To: zhbluweb at hotmail.com
>>> CC: bioconductor at stat.math.ethz.ch
>>> Subject: Re: [BioC] subset in XPS
>>>
>>> Dear Zhibin
>>>
>>> In general, xps offers two ways to get the treenames from an ExprTreeSet:
>>>
>>> 1. method treeNames applied to the ExprTreeSet:
>>>
>>>> treenames <- treeNames(data.rma)
>>>> treenames
>>>>
>>> 2. function getTreeNames applied to the root file directly:
>>>
>>>> treenames <- getTreeNames("Test3RMA.root")
>>>> treenames
>>>> treenames <- getTreeNames("Test3RMA.root","mdp")
>>>> treenames
>>>>
>>> Then you can select the treenames of interest by doing:
>>>
>>>> treenames <- treenames[c(2,4)]
>>>>
>>> Of course, the following also works:
>>>
>>>> value <- exprs(data.rma)
>>>> treenames <- colnames(value)
>>>> treenames
>>>> treenames <- treenames[c(4,6)]
>>>>
>>> In any case you get the subset:
>>>
>>>> sub.rma <- data.rma
>>>> exprs(subset.rma, treenames) <- value
>>>> str(sub.rma)
>>>>
>>> which results in:
>>> ..@ treenames:List of 2
>>> .. ..$ : chr "TestA2.mdp"
>>> .. ..$ : chr "TestB2.mdp"
>>>
>>> Could you please send me the code you used for subsetting, which
>>> resulted in an "X" in front of treenames.
>>>
>>> Do above mentioned solutions result in the same error?
>>>
>>> Furthermore, could you give me the following information:
>>> - which version of ROOT did you install?
>>> - did you install the ROOT binary or compile from source?
>>> - did you download/install the xps binary or compile from source?
>>>
>>> P.S.:
>>> I am glad to hear that running R through Terminal on your Mac works fine.
>>>
>>> Best regards
>>> Christian
>>>
>>>
>>> Zhibin Lu wrote:
>>>
>>>> Dear Christian,
>>>>
>>>> When I tried to use str(sub.rma) to check the sub set, I found there was an extra 'X' in front of each tree name. For the example you provided, the treenames were:
>>>> .. at treenames:List of 2
>>>> .. ..$ : chr "XTestA2.mdp"
>>>> .. ..$ : chr "XTestB1.mdp"
>>>>
>>>> When I applied filters to the sub.rma, I got an error "ERROR: Could not get tree .". After I changed the treenames manually, the error was gone.
>>>>
>>>> I was running R 2.8/BioC 2.3 and xps 1.1.2 under Ubuntu 8.04.
>>>>
>>>> Regards,
>>>>
>>>> Zhibin
>>>>
>>>>
>>>>
>>
>> _________________________________________________________________
>>
>>
>>
>>
>