[BioC] question regarding genefilter package

James W. MacDonald jmacdon at med.umich.edu
Tue Feb 2 21:16:33 CET 2010


zrl wrote:
> Thank you Jim. I think you pretty much answered my questions. So, the
> "var.cutoff" is actually the quantile of the IQRs calculated of all
> the rows/genes. Do I understand this right?

Yes.


> 
> Thanks.
> 
> Qiudao.
> 
> On Tue, Feb 2, 2010 at 2:28 PM, James W. MacDonald
> <jmacdon at med.umich.edu> wrote:
>> Hi Whomever,
>>
>> zrl wrote:
>>> Dear list,
>>> I am bit confused about the meaning of "var.cutoff" in function of
>>> "nsFilter". Since we calculate the IQR of each gene, how could the
>>> var.cutoff is a absolute value, 0.5.
>>> When I read the help document, it says "Note that by default the
>>> numerical-filter cutoff is interpreted as a quantile..." Does it mean
>>> we really compare the IQR with region of quantile 0 to quantile 0.5?
>>> Could someone help me clarify this? Thanks.
>> I'm not sure what you are asking, so I'll just take a stab at it.
>>
>> The 'stock' setting of nsFilter takes each row of the data in your
>> ExpressionSet and computes the IQR for that row. It then calculates what the
>> median IQR is, and then removes the 50% of the data that have an IQR below
>> that value.
>>
>> So if you change the var.cutoff to 0.2, you would remove the data with an
>> IQR below the first quintile. Or if you changed var.cutoff to 0.75, you
>> would remove the data with an IQR below the third quartile.
>>
>> Best,
>>
>> Jim
>>
>>
>>> _______________________________________________
>>> 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
>> --
>> James W. MacDonald, M.S.
>> Biostatistician
>> Douglas Lab
>> University of Michigan
>> Department of Human Genetics
>> 5912 Buhl
>> 1241 E. Catherine St.
>> Ann Arbor MI 48109-5618
>> 734-615-7826
>> **********************************************************
>> Electronic Mail is not secure, may not be read every day, and should not be
>> used for urgent or sensitive issues
>>

-- 
James W. MacDonald, M.S.
Biostatistician
Douglas Lab
University of Michigan
Department of Human Genetics
5912 Buhl
1241 E. Catherine St.
Ann Arbor MI 48109-5618
734-615-7826
**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues 



More information about the Bioconductor mailing list