[BioC] timecourse package - Missing data

[Yu Chuan Tai]

Hi Marcelo,

The current version of Timecourse is for longitudinal time course data 
only. Based on what you described below, your study looked like a 
cross-sectional one because time point samples were from different 
animals. That's, they were independent. I'd suggest that you use other 
software for analyzing your data, e.g. limma.
The cross-sectional method will be built in timecourse in the future.

Best,
Yu Chuan

On Mon, 8 Feb 2010, Marcelo Laia wrote:

>> From the timecourse vignete:
>
> "For version 1.0.0, missing time point samples are not allowed
> inmb.long() andmb.MANOVA(),
> while missing replicates are allowed for a subset of genes."
>
> Is this missing data only allowed for a subset of genes or for a whole
> replicate?
>
> We have a data set with 5 times points:
>
> time0 - animal 1, animal 2, and animal 3
> time1 - animal 4, animal 5, and animal 6
> time2 - animal 7, and animal 8
> time3 - animal 9, animal 10, animal 11
> time4 - animal 12, animal 13, animal 14
>
> So, time 2 has only 2 replicates.
>
> I try this code:
>
>> assay <- c("d1", "d1", "d1", "d1", "d2", "d2", "d2", "d2", "d2",
> "d3", "d3", "d3", "d3", "d3")
>
>> time.grp <- c(1, 2, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5)
>> time.grp
>
>> dim(todos.timecourse.matrix)
> [1] 9600   14
>
>> size <- read.table("reps.csv", sep="\t", header=TRUE)
>> size.matrix = size[,2:6]
>> row.names(size.matrix) = size[,1]
>
>> size.matrix[1:10,]
>          T0 T1 T2 T3 T4
> gene1  3  3  2  3   3
> gene2  3  3  2  3   3
> gene3  3  3  2  3   3
> gene4  3  3  2  3   3
> gene5  3  3  2  3   3
> gene6  3  3  2  3   3
> gene7  3  3  2  3   3
> gene8  3  3  2  3   3
> gene9  3  3  2  3   3
> gene10  3  3  2  3   3
>>
>
>> out1 <- mb.long(todos.timecourse.matrix, times = 5, reps = size.matrix,
> +                 rep.grp = assay, time.grp = time.grp,
> HotellingT2.only = FALSE)
> Error in mb.1D(object, times, reps, prior.df, prior.COV, prior.eta,
> time.grp,  :
>  The sample sizes are incorrect.
>>
>
>
> Are there a way to analysis this data on timecourse package?
>
> Am I doing any mistake or misinterpretation?
>
> Thank you very much
>
> --
> Marcelo Luiz de Laia
> Brazil
> Linux user number 487797
>
> _______________________________________________
> 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
>