BiSeq. Model beta regression on resampled data.
1
0
Entering edit mode
jjalmagro • 0
@jjalmagro-12099
Last seen 4.6 years ago

Hello,

I am stuck with one of the steps required to calculated DMRs using BiSeq. I have modeled the methylation under the beta regression using the group and gender as explanatory variables. However, when I reach the step where the beta regression has to be calculated for the resampled data, I am not sure about how it should be resampled when there are more than one explanatory variable. Do I have to resample just the group variable or both of them?

Regards,
Jose

biseq methylation dmr analysis • 682 views
ADD COMMENT
0
Entering edit mode
@katja-hebestreit-6495
Last seen 12 months ago
United States
Hi Jose, The additional explanatory variable shouldn't change much for the resampling. If your explanatory variable of interest (group) has 2 values, the two sampled groups should contain the same number of samples with each value. The only thing that is different is that you also need to add gender as explanatory variable to your model (with the original values per sample, so you don't assign new values for this one), assuming you are not interested in finding differentially methylated regions in regard to gender. Cheers, Katja On Tue, Jan 3, 2017 at 2:28 AM, jjalmagro [bioc] <noreply@bioconductor.org> wrote: > Activity on a post you are following on support.bioconductor.org > > User jjalmagro <https: support.bioconductor.org="" u="" 12099=""/> wrote Question: > BiSeq. Model beta regression on resampled data. > <https: support.bioconductor.org="" p="" 90725=""/>: > > Hello, > > I am stuck with one of the steps required to calculated DMRs using BiSeq. > I have modeled the methylation under the beta regression using the group > and gender as explanatory variables. However, when I reach the step where > the beta regression has to be calculated for the resampled data, I am not > sure about how it should be resampled when there are more than one > explanatory variable. Do I have to resample just the group variable or both > of them? > > Regards, > Jose > > ------------------------------ > > Post tags: biseq, methylation, dmr analysis > > You may reply via email or visit BiSeq. Model beta regression on resampled data. >
ADD COMMENT
0
Entering edit mode

Hello Katja,

I get it know, thank you. One final question, I have modeled the methylation using 12 cases and 6 controls. Do I have to use just 6 cases and  the 6 controls to have the same number of samples in each group?

Regards,
Jose
 

ADD REPLY
0
Entering edit mode
Hi Jose, You could also do this: group 1 with 6 cases and 3 controls and group 2 with 6 cases and 3 controls. As long as you have the same ratio of controls to cases in both groups you should be fine. It's actually better to include a similar number of samples in your resampled data as you use in the original analysis. Cheers, Katja On Tue, Jan 3, 2017 at 11:58 AM, jjalmagro [bioc] <noreply@bioconductor.org> wrote: > Activity on a post you are following on support.bioconductor.org > > User jjalmagro <https: support.bioconductor.org="" u="" 12099=""/> wrote Comment: > BiSeq. Model beta regression on resampled data. > <https: support.bioconductor.org="" p="" 90725="" #90752="">: > > Hello Katja, > > I get it know, thank you. One final question, I have modeled the > methylation using 12 cases and 6 controls. Do I have to use just 6 cases > and the 6 controls to have the same number of samples in each group? > > Regards, > Jose > > > ------------------------------ > > Post tags: biseq, methylation, dmr analysis > > You may reply via email or visit https://support.bioconductor. > org/p/90725/#90752 >
ADD REPLY

Login before adding your answer.

Traffic: 666 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6