[BioC] MA plot (using fitted M values)

[Choon Wei Wee]

This is the first time I am asking question on Bioconductor and I am not sure which address I should be emailing to. Hope I didn't send to the wrong address.

Hi

I am using LimmaGUI at the moment to analyse quite a large dataset which includes lucidea spiking controls. I realise that when I used LimmaGUI to plot MA plot on individual slide with coloured legend, everything is fine. However, when I made MA plot using linear modeled data (on all 4 slides), there seemed to be colour distortion, i.e., the ratio controls will have the same colour as my genes etc. Do anyone have any ideas to as how I can get around with it?

Also, as I am quite a new player with working with array data, I will certainly appreciate if anyone can offer me any ideas to as what kind of statistical tests should I do with array data if I am preparing them for publications. Right now, I am just using the calibration controls (none DE genes) in lucidea scorecard to arbitrary set cut offs for DE genes. In most cases, the genes I termed as differentially expressed will have good B, T and P values from LimmaGUI and some of the genes had been verified to be DE. But are these (statistics) enough for publication?

Thank you.

Choon Wei Wee
Centre for Environmental Stress and Adaptation Research, Dept of Genetics, University of Melbourne, Australia.
	[[alternative HTML version deleted]]