Hi- In the roast paper, the examples tended to have extra samples not used to define the contrast - which helped increase the degrees of freedom for the rotation test (e.g. 26 samples in a 3 group experiment; group 1 has 3 samples, group 2 has 3 samples and group 3 has 20 samples. - the contrast compared group 1 and group 2). Is running roast on a dataset with 6 samples in 2 groups (3 test, 3 control) an acceptable use case or will the degrees of freedom be too small to get valid p-values? Thanks, Julie Julie Leonard Computational Biologist Global Bioinformatics Syngenta Biotechnology, Inc. 3054 E. Cornwallis Rd. Research Triangle Park, NC 27709 USA julie.leonard@syngenta.com<mailto:julie.leonard@syngenta.com> www.syngenta.com<http: www.syngenta.com=""/> This message may contain confidential information. If yo...{{dropped:7}}

Hi Julie, Thank you for asking. What you described is an acceptable case. ROAST has increased power when you have extra samples that are not in the contrast but in the dataset, however it's not required to have those extra samples. P value is still valid. Hope this help. Let me know if you have further questions. Cheers, Di ---- Di Wu Postdoctoral fellow Harvard University, Statistics Department Harvard Medical School Science Center, 1 Oxford Street, Cambridge, MA 02138-2901 USA ________________________________________ From: bioconductor-bounces@r-project.org [bioconductor- bounces@r-project.org] on behalf of julie.leonard@syngenta.com [julie.leonard@syngenta.com] Sent: Tuesday, June 25, 2013 4:38 PM To: bioconductor at r-project.org Subject: [BioC] running roast with 6 samples in 2 groups Hi- In the roast paper, the examples tended to have extra samples not used to define the contrast - which helped increase the degrees of freedom for the rotation test (e.g. 26 samples in a 3 group experiment; group 1 has 3 samples, group 2 has 3 samples and group 3 has 20 samples. - the contrast compared group 1 and group 2). Is running roast on a dataset with 6 samples in 2 groups (3 test, 3 control) an acceptable use case or will the degrees of freedom be too small to get valid p-values? Thanks, Julie Julie Leonard Computational Biologist Global Bioinformatics Syngenta Biotechnology, Inc. 3054 E. Cornwallis Rd. Research Triangle Park, NC 27709 USA julie.leonard at syngenta.com<mailto:julie.leonard at="" syngenta.com=""> www.syngenta.com<http: www.syngenta.com=""/> This message may contain confidential information. If yo...{{dropped:7}} _______________________________________________ Bioconductor mailing list Bioconductor at r-project.org https://stat.ethz.ch/mailman/listinfo/bioconductor Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor