[BioC] BioC 1.4 MAS implementation

[James W. MacDonald]

Lizhe Xu wrote:
> Please help me with the MAS method. 
> I got more than half of the probe sets with value as NA (16645 for U133A) after using the following command:
> 
> 
>>set<-expresso(data, normalize.method="constant", bgcorrect.method="mas", pmcorrect.method="subtractmm",summary.method="mas")
> 
> background correction: mas 
> normalization: constant 
> PM/MM correction : subtractmm 
> expression values: mas 
> background correcting...done.
> normalizing...done.
> 22283 ids to be processed
> |                    |
> |####################|
> There were 50 or more warnings (use warnings() to see the first 50)
> 
> When I checked the warnings it showed
> 
>>warnings()
> 
> Warning messages:
> 1: NaNs produced in: log(x, base) 
> 2: NaNs produced in: log(x, base) 
> 3: NaNs produced in: log(x, base) 
> 4: NaNs produced in: log(x, base) 
> 5: NaNs produced in: log(x, base) 
> 6: NaNs produced in: log(x, base) 
> 7: NaNs produced in: log(x, base) 
> 8: NaNs produced in: log(x, base) 
> 9: NaNs produced in: log(x, base) 
> 10: NaNs produced in: log(x, base) 
> 
> Did I make some mistake on the command? What do the warnings mean?

These warnings indicate that you are trying to take the logarithm of a 
negative number, which is one of the downfalls of the original MAS4.0 
method. Are you trying to emulate the MAS5.0 algorithm, or do you really 
want to do what you have here?

For MAS5.0 values, you can simply try

set <- mas5(data) ## note here that 'data' is not a good variable name 
because you are masking an existing R function. You might substitute dat 
or Data.

If you really want to do what you have in your call to expresso(), then 
you will likely have to change your pmcorrect.method, because this will 
result in many negative values. Since the first step in the 'mas' 
summary stat calculation is to take logs, this will always be a problem.

HTH,

Jim


> 
> Thanks.
> 
> 
> L
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor


-- 
James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109