Develop

[Cybernewbye]

[Dear Dr. Peschel

Thank you for your great job and the easiness of use of your code, which is quite uncommon. I'm quite unaware of network science, and I'd like to apply it on our metabarcoding analysis as a follow up in environmental bioremediation applications. The question that we are unable to address is how to handle longitudinal data (e.g. five monthly sampling during 6 months test, that we would like to consider as non-independent). I was wondering if any of the already coded tools that you greatly provided is able to handle this kind of relationship between samples (and replicates), or if it would be possible to introduce other modeling methods like ARZIMM (that for a profane like me, produces a very long list of numbers that I could not use to draw any network). sorry for any improper term or lack of knowledge that might emerge in this message

thank you for your kind attention

[stefpeschel]

Hey @Cybernewbye,

Thanks for your interest in using NetCoMi for your analyses. However, the package is designed to compare two networks only, which can be two time points, environmental conditions, etc. NetCoMi does not allow you to compare - and especially test - more than two networks. You would only be able to compare multiple networks graphically, using the same layout for each time point. We have further functionality, e.g. for multi-domain analysis, in the pipeline, but implementing ARZIMM is not planned.

Regarding your pull request: The develop branch should currently not be merged into main. So, I will not accept your request. Please fork the package if you want to modify the code yourself.

Best, Stefanie