PCoA

naman · 06-26-2015, 03:01 AM

Dear All,

I have 16s data and have some queries regarding the same.
I am doing PCoA analysis and Axis 1 and 2 are not showing a separation of sample based on specific factor, so i plotted to axis 3 and 4 which shows sample separation.
My question is it unfair to choose axis other than 1 and 2 which generally show maximum variance.
I always assumed that Statistical test will tell you the significant differences and then you can plot any 2 axis which shows the separation of significant factors.
Sometimes first two axis are not always the correct axis showing the variance for the factor which are found significant p value.
And that's the reason i changed the axis.
Please provide me a valuable suggestion to undo my dilemma
Also if someone have a scientific journal where they too have used other that first and second axis.
Thank you looking forward for the reply!!

06-26-2015, 03:01 AM	#1
naman Junior Member Location: Germany Join Date: Sep 2012 Posts: 8	PC Axis? Dear All, I have 16s data and have some queries regarding the same. I am doing PCoA analysis and Axis 1 and 2 are not showing a separation of sample based on specific factor, so i plotted to axis 3 and 4 which shows sample separation. My question is it unfair to choose axis other than 1 and 2 which generally show maximum variance. I always assumed that Statistical test will tell you the significant differences and then you can plot any 2 axis which shows the separation of significant factors. Sometimes first two axis are not always the correct axis showing the variance for the factor which are found significant p value. And that's the reason i changed the axis. Please provide me a valuable suggestion to undo my dilemma Also if someone have a scientific journal where they too have used other that first and second axis. Thank you looking forward for the reply!! Last edited by naman; 10-01-2015 at 05:10 AM.