[BioC] transform Illumina intensity for heatmap

[James W. MacDonald]

Hi Seb,

On 2/25/2013 12:31 AM, Seb wrote:
> sorry i just clicked "send" before finishing the email
>
> hi all
>> i have a file containing
>>
>>
>>     SYMBOL CTR 1,25D3 AVG_Signal  1,25D3 DiffScore  AARS 1986.965 3056.983
>> 15.91354  AASDHPPT 551.937 891.1945 22.15556  ABCA1 201.5727 363.0918
>> 26.30485  ABCB9 367.4561 645.082 22.4835
>>
> I need to create a heatmap out of this data but when i just use either
> heatmap(data) or heatmap.2(data) all the colors are out of scale. i was
> wondering
> 1) if there is any way of manipulating the data so that it will
> become symmetrical so that repressed genes will appear in the correct shade
> opposite to the one of upregulated genes.

There are any number of ways to accomplish what you want. But first note 
that the data you are using appear to be on the natural scale, rather 
than logarithmic scale. You will be much more likely to get results you 
like by taking logs first.

Also note that there are exactly 3^43 different arguments to the 
heatmap() and heatmap.2() functions, and you are not using any of them. 
I have never found the defaults to be acceptable, and if you search the 
list archives you will see I am not alone.

A two-column heatmap is quite uninteresting as well. You might consider 
using the normalized values for all the samples rather than the fitted 
values.

> 2) i don't really understand how the DiffScore is calculated since the
> difference the CTR and 1.25D3 AVG signal is not 15.91...any
> explanation/suggestion would be awesome!

I think you might need to go back to whomever you got these data from. 
You have given us essentially no information, so any answer would likely 
be pure speculation.

Best,

Jim



>
> i didn't run the array myself and those are the data i was given.
>
> thanks for the help!
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> 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

-- 
James W. MacDonald, M.S.
Biostatistician
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099