[BioC] which statistical test to perform?

[Sean Davis]

Hi, Anand.

Please try to keep the conversations on the list so that you can get
the best answers to your questions.

First, you will need to split the data back out to include the values
for all your replicates.  In other words, do not use the means.
Working with means only essentially precludes any statistical testing
at all.

Second, I would suggest that you take a look at the limma package and
the wonderful limma user guide.  The statistical framework used in
limma is the linear model and it works well for two-class or
multi-class problems.

Sean


On Wed, Nov 23, 2011 at 6:40 AM, anand m t <anandrox05 at gmail.com> wrote:
> Sir,
> I'm tying to compare the data from all the brain tissues. The data which
> i've shown here is the mean value of all 3 biological replicates of each
> tissue.
>
> On Wed, Nov 23, 2011 at 4:57 PM, Sean Davis <sdavis2 at mail.nih.gov> wrote:
>>
>> On Wed, Nov 23, 2011 at 5:14 AM, anand mt [guest]
>> <guest at bioconductor.org> wrote:
>> >
>> > hi all,
>> >
>> > i'm new to this microarray data  analysis.
>> > recently i've been given data consisting of 11 tissues.
>> > now i've normalized the data, filtered data using mas5 AP calls. My
>> > question is, which statistical test i must perform
>> > to calculate the significance values ??
>> >
>> > sample data is as below:
>> >
>> >
>> >        accumbens       amygdala        cerebellum      corpus.collosum
>> > hippocampus     midbrain        p.lobe  putamen s.nigra t.lobe  thalamus
>> > 1007_s_at       11.93852233     12.21404093     11.46118612
>> > 13.41594885     12.42216256     12.89589133     11.58715914     11.85803472
>> >     12.79920479     12.07087932     12.55338306
>> > 1053_at 7.490706858     7.526181155     7.551069308     7.891002293
>> > 7.49104271      7.971097552     8.088918072        7.660258014
>> > 7.92423132      7.54689645      7.128753703
>> > 117_at  8.486898268        8.773089087        7.642339349
>> > 8.560352732     7.676296801     7.865961146        7.250275943
>> > 7.929165261     7.874073766        7.940298941     8.10731601
>> >
>> >
>> > I got some web results, from which i came to know that, chi-square test
>> > is of more relevant in this case (to compare 3 or more unmatched groups,
>> > binomial). Is it correct to choose chi-square test ??
>> >
>> > Sorry if my question is too lame.
>>
>> Hi, Anand.
>>
>> I'm assuming that for you the biological question that you are asking
>> is obvious, but to me it seems unclear.  In particular, what groups
>> above are you trying to compare?  It seems you have no replicates?
>>
>> Sean
>>
>>
>> > thanks in advance.
>> >
>> >  -- output of sessionInfo():
>> >
>> > R version 2.13.1 (2011-07-08)
>> > Platform: i386-pc-mingw32/i386 (32-bit)
>> >
>> > locale:
>> > [1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United
>> > States.1252    LC_MONETARY=English_United States.1252 LC_NUMERIC=C
>> > [5] LC_TIME=English_United States.1252
>> >
>> > attached base packages:
>> > [1] stats     graphics  grDevices utils     datasets  methods   base
>> >
>> > other attached packages:
>> > [1] MASS_7.3-13
>> >
>> > loaded via a namespace (and not attached):
>> > [1] tools_2.13.1
>> >
>> > --
>> > Sent via the guest posting facility at bioconductor.org.
>> >
>> > _______________________________________________
>> > Bioconductor mailing list
>> > Bioconductor at r-project.org
>> > https://stat.ethz.ch/mailman/listinfo/bioconductor
>> > Search the archives:
>> > http://news.gmane.org/gmane.science.biology.informatics.conductor
>> >
>
>
>
> --
> ******************************************************************
> Anand M.T
> School of Biotechnology (Bio-Informatics),
> International Instituteof Information Technology (I2IT),
> P-14, Rajiv Gandhi Infotech park,
> Hinjewadi,
> Pune-411 057.
> INDIA.
>