Corr.axes metadata missing values

I’m trying to run the corr.axes command in mothur v.1.43.0 (latest update: 10/21/19) with the metadata option as described in the mothur sop.

corr.axes(, metadata=XXmetadata.metadata, method=spearman, numaxes=3)

Axes File:
group axis1 axis2 axis3 axis4 axis5 axis6
A01a -0.403967 -0.041514 0.028161 -0.000499 -0.024524 0.013728
A01b -0.383460 0.012639 -0.013652 0.007365 0.007631 -0.010989
A02a -0.392569 -0.003826 -0.007022 0.001669 0.009751 -0.006197
A02b -0.398429 -0.014096 -0.002861 -0.009460 0.011359 -0.006334
RB15c 0.377365 -0.004237 -0.099474 -0.030461 -0.149048 -0.217283
RC01a 0.316966 0.125874 -0.013639 0.038477 0.090080 0.026957

Metadata File:
group fr ss n c a
A01a S A P w o
A01b S A P w o
A02a S A F d o
A02b S A F d o
RB15c R B L d n
RC01a R C P w o

Unfortunately I get an empty file:
Feature axis1 p-value axis2 p-value axis3 p-value length

Within Mothur it is said:
"You did not provide a label, I will use the first label in your inputfile.

Output File Names:

I modified the metadata file several times to try with different numbers of features and also put quotation marks around describing features, but none of my modifications worked so far.

May someone is able to explain what might be wrong!!


Hi Deel,

The metadata file needs to contain numeric data, not categorical. If you want to compare categorical data, then you’d likely want to use something like amova or homova; however, with only 6 samples I’m not sure you will have much statistical power to say much.


1 Like

Hi Pat,

Thank you very much for helping me this fast!!

I converted my categorical data into a numeric scheme to destinguish groups.
Now the corr.axes command works perfectly fine!

Of course only 6 samples have no statistical power. I just wanted you to have an idea how my data files look like to make it easier helping me. But thank you very much for this hint!! I really appreciate that!

Best regards,

This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.