hello
I have only OTU tables for three metagenomic samples ( three groups each one contains one sample) can I run lefse on this small sample size? is it possible to be statistically effective?
thank you
1 Like
Hi there -
I doubt so. In this case you won’t have any samples to estimate the within-group variability with. I’d say it’s likely that the method won’t return a meaningful p-value (NA or 1).
Thanks,
Siyuan
Hello Siyuan and everyone,
I have a question that might be related to this topic: I have been using Lefse to identify differences between microbial communities (each containing several samples), so far I have obtained nice results for these analyses.
I was now interested in comparing a single sample, acting as group one, to a second group (composed of multiple samples). I was wondering if LEfSe was designed for this type of scenarios.
In case LEfSe wasn’t intended for this purpose I would really appreciate if you could point me into the right direction. What other program/approach I could use to perform this type of comparisons?
Best Regards,
David
Hi David,
It is unlikely that comparing one sample to a group of samples will tell you much, especially if these are human samples. Interpersonal differences are so high in human-related samples that this type of comparison would simply tell you what is unique about that person’s microbiome, but also would not return meaningful p-values as Siyuan indicated above.
In environmental samples- the power is going to be a bigger problem. Similar to what Siyuan said above it won’t be able to estimate the within-group variability. That said I don’t know how you could do this comparison with statistical power. Sorry that I cannot provide a better answer.
Best, Kelsey
Hello Kelsey,
thank you for your response, what you are saying makes total sense. It would be hard to extract a valid conclusion from such comparison.
Best Regards,
David