S3 support for phyloseq and TreeSE

[antagomir]

Opening this PR related to philr issue #16

All checks pass and this can be reviewed & necessary changes discussed but two things are pending before merge:

• Confirmation that the mia package Bioc devel version has been updated (working on this) as some functionality for TreeSE examples is required from there
• Links in the vignette (between sub-vignettes) may need an update, could not check fully yet

Kindly review and suggest any modifications.

[antagomir]

This PR is now ready.

Summary of changes:

• Added support for TreeSummarizedExperiment and phyloseq in the philr function via S3 classes

• Revised the vignette to include examples with and without these data formats within a single workflow; we maintain the same workflow than previously; the independent phyloseq and miaverse frameworks can provide additional capabilities for tree handling, ordination, visualization etc. but I aimed to keep things simple and focused on philr functionality in this vignette.

• Some minor changes added that were requested by Bioconductor checks

After merging, this should be bumped to Bioc devel.

[jsilve24]

Thanks for putting all this together. I think this is a very nice contribution. Overall I just had two comments about adding required dependencies and about changing the default behavior of the philr function.

Happy to discuss more if you feel strongly on either.

[antagomir]

All requested changes implemented, with some bioc stilization. Happy to make further fixes if there is a need.

We could change philr input argument df to x on the same go if this is desired. Or leave as is.

[jsilve24]

Thanks again for all this work. I think I found one error in the proposed pull -- defaults don't seem to have been changed in the main s3 function, only in the individual methods.
I think this will cause problems.

[antagomir]

I think I have now made all requested changes but if not, let us continue discussion & fixing.

[jsilve24]

This looks great. Also I think you are correct, my appologies for the confusion regarding the default weights. Thank you for your patience and hard work. Merging now.

[antagomir]

Awesome, thanks - pleasure to contribute to such a great package! The phylofactor is possible next target.

I anticipate that the updates will be also pushed to Bioconductor devel upstream to facilitate further use: I noticed that running the standard check BiocCheck::BiocCheck() generates some more suggested coding style updates. But I expect that those are not mandatory for already accepted packages.