Can anyone tell me what the units of the numbers on the Y-axis of the PCA plot are? I think its the number of genes involved in that particular component but maybe someone can tell me for sure? See here for an example of what I'm talking about: http://frink.nuigalway.ie/~pat/PCAScreePlotNorm.pdf Thanks alot, -Paul.

Y=variance of component units are measurement units^2 --Naomi At 10:05 AM 6/16/2009, you wrote: >Can anyone tell me what the units of the numbers on the Y-axis of the >PCA plot are? I think its the number of genes involved in that >particular component but maybe someone can tell me for sure? See here >for an example of what I'm talking about: > >http://frink.nuigalway.ie/~pat/PCAScreePlotNorm.pdf > >Thanks alot, > >-Paul. > >_______________________________________________ >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 Naomi S. Altman 814-865-3791 (voice) Associate Professor Dept. of Statistics 814-863-7114 (fax) Penn State University 814-865-1348 (Statistics) University Park, PA 16802-2111

The fun thing about eigenvalues is that they are linear combinations of ALL the genes. At 10:54 AM 6/16/2009, you wrote: >Oh right, is there any way to figure out how many genes are involved >in the component or does that even make sense? > >On Tue, Jun 16, 2009 at 3:39 PM, Naomi Altman<naomi at="" stat.psu.edu=""> wrote: > > Y=variance of component > > > > units are measurement units^2 > > > > --Naomi > > > > At 10:05 AM 6/16/2009, you wrote: > >> > >> Can anyone tell me what the units of the numbers on the Y-axis of the > >> PCA plot are? I think its the number of genes involved in that > >> particular component but maybe someone can tell me for sure? See here > >> for an example of what I'm talking about: > >> > >> http://frink.nuigalway.ie/~pat/PCAScreePlotNorm.pdf > >> > >> Thanks alot, > >> > >> -Paul. > >> > >> _______________________________________________ > >> 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 > > > > Naomi S. Altman 814-865-3791 (voice) > > Associate Professor > > Dept. of Statistics 814-863-7114 (fax) > > Penn State University 814-865-1348 (Statistics) > > University Park, PA 16802-2111 > > > > Naomi S. Altman 814-865-3791 (voice) Associate Professor Dept. of Statistics 814-863-7114 (fax) Penn State University 814-865-1348 (Statistics) University Park, PA 16802-2111

Hi Paul When I've done this before I have examined the scree plot to see which components are contributing most variance. Say PC1 and PC2 both show a considerable contribution. I then plot these against each other and look to see if there are points high  / low in one PC but not in another. Finally I pull out the loadings for these PCs and use that to identify which genes are 'different' in the two components. I usually do this last step in a speadsheet (Excel, OOO Calc) by placing the loadings into two columns and then ordering these columns by one of them (usually PC1) and then looking at the value differences across both columns. I'm not entirely sure this is the right way to go but it's given me useful information in the past. Cheers Iain --- On Tue, 16/6/09, Paul Geeleher <paulgeeleher@gmail.com> wrote: From: Paul Geeleher <paulgeeleher@gmail.com> Subject: Re: [BioC] PCA scree plot question To: "Naomi Altman" <naomi@stat.psu.edu> Cc: bioconductor@stat.math.ethz.ch Date: Tuesday, 16 June, 2009, 5:52 PM So it doesn't make sense to say that subset of genes is exclusively responsible for a single component? I take what you are saying to mean that every gene contributes to every component but to a different degree. I'll have to do some background reading on this PCA stuff... On Tue, Jun 16, 2009 at 4:34 PM, Naomi Altman<naomi@stat.psu.edu> wrote: > The fun thing about eigenvalues is that they are linear combinations of ALL > the genes. > > At 10:54 AM 6/16/2009, you wrote: >> >> Oh right, is there any way to figure out how many genes are involved >> in the component or does that even make sense? >> >> On Tue, Jun 16, 2009 at 3:39 PM, Naomi Altman<naomi@stat.psu.edu> wrote: >> > Y=variance of component >> > >> > units are measurement units^2 >> > >> > --Naomi >> > >> > At 10:05 AM 6/16/2009, you wrote: >> >> >> >> Can anyone tell me what the units of the numbers on the Y-axis of the >> >> PCA plot are? I think its the number of genes involved in that >> >> particular component but maybe someone can tell me for sure? See here >> >> for an example of what I'm talking about: >> >> >> >> http://frink.nuigalway.ie/~pat/PCAScreePlotNorm.pdf >> >> >> >> Thanks alot, >> >> >> >> -Paul. >> >> >> >> _______________________________________________ >> >> 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 >> > >> > Naomi S. Altman                                814-865-3791 (voice) >> > Associate Professor >> > Dept. of Statistics                              814-863-7114 (fax) >> > Penn State University                         814-865-1348 (Statistics) >> > University Park, PA 16802-2111 >> > >> > > > Naomi S. Altman                                814-865-3791 (voice) > Associate Professor > Dept. of Statistics                              814-863-7114 (fax) > Penn State University                         814-865-1348 (Statistics) > University Park, PA 16802-2111 > > _______________________________________________ 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]]