Hi there, I am trying to use LIMMA to analyze gene expression data from an experiment which has dose response but only one replicate at each dose. I tried to fit a linear model using lmfit(). I used the doses as continuous variable. I do the ebayes fit and finally do decide tests with adj p values (BH corrected). I get coefficients, intercept and dose as output with t-stat and p values for each. I was wondering how to interpret these. What does intercept, dose and coefficients mean in this case? The data matrix I read into R was Affy Plus2 chip data for 4 doses of a compound. Any help would be greatly appreciated. Best Regards,Som. [[alternative HTML version deleted]]

Hello Somnath: Can you please post some example of the data and the code? It will make help to make your description clearer. If I understand correctly, you are assuming the replicates to be identical; so if rep_1 receives dose_1 and rep_2 receives dose_2, there is no replicate effect for your response (equivalent to assuming no random rep effect). basically your data is like: dose1 dose2 r11 r12 ..... r21 r22 .... g1 x x x x g2 x x x x . . rij represent the jth replicate for ith treatment If I am correct, the adj. p-values are for the hypotheses tests for differential gene expression corrected for multiple comparisons. The coefficients are the log fold change for a particular gene if dose changes from 1 -> 2. All these interpretations hold under the assumption that replicates are "identical". Thanks, S. On Tue, Nov 9, 2010 at 7:17 PM, somnath bandyopadhyay < genome1976@hotmail.com> wrote: > > Hi there, > I am trying to use LIMMA to analyze gene expression data from an experiment > which has dose response but only one replicate at each dose. I tried to fit > a linear model using lmfit(). I used the doses as continuous variable. I do > the ebayes fit and finally do decide tests with adj p values (BH corrected). > I get coefficients, intercept and dose as output with t-stat and p values > for each. I was wondering how to interpret these. What does intercept, dose > and coefficients mean in this case? The data matrix I read into R was Affy > Plus2 chip data for 4 doses of a compound. Any help would be greatly > appreciated. > Best Regards,Som. > [[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]]

Without the model it is hard to say, but I am guessing you fitted a straight line for each gene. So the intercept is the intercept of the line, and the coefficient for dose is the slope. The t-stat is the test of whether the intercept and slope are equal zero. The intercept is not interesting for your purpose. The slope tells you if there is a statistically significant dose effect. Regards, Naomi At 07:17 PM 11/9/2010, somnath bandyopadhyay wrote: >Hi there, >I am trying to use LIMMA to analyze gene expression data from an >experiment which has dose response but only one replicate at each >dose. I tried to fit a linear model using lmfit(). I used the doses >as continuous variable. I do the ebayes fit and finally do decide >tests with adj p values (BH corrected). I get coefficients, >intercept and dose as output with t-stat and p values for each. I >was wondering how to interpret these. What does intercept, dose and >coefficients mean in this case? The data matrix I read into R was >Affy Plus2 chip data for 4 doses of a compound. Any help would be >greatly appreciated. >Best Regards,Som. > [[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