meta-analysis with P values
2
0
Entering edit mode
@laura-rodriguez-murillo-3129
Last seen 9.6 years ago
Hi dear list, My question is on meta-analysis. I have three studies on the association of a SNP with a disease and I need to do a random-effects meta-analysis with my data. I only have the individual P value, the odds ratio and sample size from each study. I've checked rmeta and meta packages but it seems to me that I need to know the number of observations on each group, which I don't have in this case. Is there any other package I can use that accepts doing meta-analysis with these data? Thanks a lot for your help! Laura [[alternative HTML version deleted]]
SNP SNP • 3.1k views
ADD COMMENT
0
Entering edit mode
@vincent-j-carey-jr-4
Last seen 16 days ago
United States
This is really a question for R-help. However, rmeta could be used if you knew the study effect sizes and standard errors (see meta.summaries). Under certain assumptions you would be able to derive a standard error for a given estimate from a p-value and sample size. On Wed, Jul 14, 2010 at 3:55 PM, Laura Rodriguez Murillo <laura.lmurillo at="" gmail.com=""> wrote: > Hi dear list, > > My question is on meta-analysis. I have three studies on the association of > a SNP with a disease and I need to do a random-effects meta-analysis with my > data. I only have the individual P value, the odds ratio and sample size > from each study. I've checked rmeta and meta packages but it seems to me > that I need to know the number of observations on each group, which I don't > have in this case. Is there any other package I can use that accepts doing > meta-analysis with these data? > > Thanks a lot for your help! > > Laura > > ? ? ? ?[[alternative HTML version deleted]] > > _______________________________________________ > 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 >
ADD COMMENT
0
Entering edit mode
I agree with Vince about posting to R-help. However, if you have individual P-values, I am not sure why you cannot do a ordinary meta-analysis combining p-values using Fisher's method. There is not an R package to do this, but it would simply involve calculating T = -2*log(p-value1)-2*log(p-value2)-...-2*log(p-valuem) where m is the number of p-values. Under the null hypothesis, this will be distributed as chi-squared with m degrees of freedom, so you would then do T.pval = 1-pchisq(T,m) to get the combined p-value. HTH, Debashis On Wed, Jul 14, 2010 at 11:01 PM, Vincent Carey <stvjc at="" channing.harvard.edu=""> wrote: > This is really a question for R-help. ?However, rmeta could be used if > you knew the study effect sizes and standard errors (see > meta.summaries). ?Under certain assumptions you would be able to > derive a standard error for a given estimate from a p-value and sample > size. > > On Wed, Jul 14, 2010 at 3:55 PM, Laura Rodriguez Murillo > <laura.lmurillo at="" gmail.com=""> wrote: >> Hi dear list, >> >> My question is on meta-analysis. I have three studies on the association of >> a SNP with a disease and I need to do a random-effects meta- analysis with my >> data. I only have the individual P value, the odds ratio and sample size >> from each study. I've checked rmeta and meta packages but it seems to me >> that I need to know the number of observations on each group, which I don't >> have in this case. Is there any other package I can use that accepts doing >> meta-analysis with these data? >> >> Thanks a lot for your help! >> >> Laura >> >> ? ? ? ?[[alternative HTML version deleted]] >> >> _______________________________________________ >> 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 >> > > _______________________________________________ > 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 > -- Debashis Ghosh Departments of Statistics and Public Health Sciences Penn State University University Park, PA 16802 works.bepress.com/debashis_ghosh/
ADD REPLY
0
Entering edit mode
Paul Leo ▴ 970
@paul-leo-2092
Last seen 9.6 years ago
You can use the metal program available at the broad website with what you have. If you have the standard error you can try the inverse varience option The bioconductor geneMeta program has REM specifically : uses the t-statistic which I think you can get if you have the standand error and the OR. It would require minor modification of some function however. The number of observations in each group is the number of SNPs no? Cheers Paul -----Original Message----- From: Laura Rodriguez Murillo <laura.lmurillo@gmail.com> To: bioconductor@stat.math.ethz.ch Subject: [BioC] meta-analysis with P values Date: Wed, 14 Jul 2010 15:55:26 -0400 Hi dear list, My question is on meta-analysis. I have three studies on the association of a SNP with a disease and I need to do a random-effects meta-analysis with my data. I only have the individual P value, the odds ratio and sample size from each study. I've checked rmeta and meta packages but it seems to me that I need to know the number of observations on each group, which I don't have in this case. Is there any other package I can use that accepts doing meta-analysis with these data? Thanks a lot for your help! Laura [[alternative HTML version deleted]] _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/bioconductor Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor [[alternative HTML version deleted]]
ADD COMMENT

Login before adding your answer.

Traffic: 837 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