[BioC] Robustspline and two dimentional loess for two color agilent data

Gordon K Smyth smyth at wehi.EDU.AU
Fri Aug 22 08:26:03 CEST 2014

Dear Samane,

You can use robustspline as an alternative to loess for any two-colour 

Robustspline should be better than print-tip-loess when there are multiple 
print-tip groups.

When there are no print-tip groups, robustspline and loess should be 
similar.  However robustspline is more robust than loess.  Technically, it 
has a higher breakdown point than loess.  This means that it can tolerate 
a greater proportion of outliers or DE probes than loess.

Robustspline has been relatively little used and cited in the literature 
because I have not published the method or actively promoted it.  However 
this is just a reflection of my lack of time, not that the method doesn't 
work well.  If you find that it works for your data, there is no reason 
why you shouldn't use it.

Best wishes

On Fri, 22 Aug 2014, samane fazeli wrote:

> Dear Dr. Smyth,
> Thank you for your reply. I am comparing within-array normalization
> methods (Median, Loess, robust-spline) in Agilent data. In order to
> select the best method in my data, I computed mean of variability
> among replicated arrays. Based on this criteria robust-spline
> outperforms loess method. This difference was significant (Wilcoxon
> test, p-value<0.05). In this situation, Can I apply robust-spline
> method on agilent data.
> Regards
> Samane
> On 8/22/14, Gordon K Smyth <smyth at wehi.edu.au> wrote:
>> Dear Samaneh Fazeli,
>> I am guessing that this is question about
>>   normalizeWithinArrays(RG, method="robustspline")
>> in the limma package.  When there are no print tips, robust-spline
>> normalization reduces to loess normalization.  Hence you should use:
>>   normalizeWithinArrays(RG, method="loess")
>> for Agilent arrays.
>> Best wishes
>> Gordon
>>> Date: Thu, 21 Aug 2014 01:54:20 -0700 (PDT)
>>> From: "Samane [guest]" <guest at bioconductor.org>
>>> To: bioconductor at r-project.org, samanefazeli at gmail.com
>>> Subject: [BioC] Robustspline and two dimentional loess for two color
>>> 	agilent	data
>>> Hi
>>> Since there is no print tip in Agilent technology, Could I use
>>> robust-spline and two dimensional loess on such data sets? I am
>>> comparing some normalization methods on two color Agilent data; In the
>>> most of times, robust-spline goes the best method based on comparison of
>>> variance and ICC among replicated arrays. However I can not find lots of
>>> papers which have applied these methods on Agilent data.
>>> I am looking forward to hearing from u.
>>> Regards,
>>> Samaneh Fazeli
>>> -- output of sessionInfo():
>>> R
>>> --
>>> Sent via the guest posting facility at bioconductor.org.
>> ______________________________________________________________________
>> The information in this email is confidential and intended solely for the
>> addressee.
>> You must not disclose, forward, print or use it without the permission of
>> the sender.
>> ______________________________________________________________________

The information in this email is confidential and intend...{{dropped:4}}

More information about the Bioconductor mailing list