Closed msmcfarlin closed 2 years ago
Hi @msmcfarlin -- unfortunately we looked into this a few years ago and doing this properly was computationally infeasible. At this time it is triaged; sorry. I won't close your issue, but just so you know, we won't be prioritizing this at the moment.
Thanks for understanding -- there's so many cool stats things that we're working on instead!
Out of curiosity, did you find the unifrac distance calculation itself was the bottleneck? (Say as compared to other beta diversity measures like Euclidean or bray Curtis?)
Actually, we decided that if we wanted to truly estimate the uncertainty in e.g., UniFrac, we should account for the uncertainty in estimating the tree as well... and that was the bottleneck. I don't like underestimating variance, so we decided to leave this alone and focus our phylogenetic uncertainty investigations elsewhere -- @svteichman is making amazing progress on this! (but it won't look like DivNet) Happy to answer more questions @mooreryan , and @paulinetrinh can chime in too if there's anything I've forgotten.
Thanks @adw96 for update! I look forward to following the work from you guys, these are some very interesting and useful packages.
Closing - as mentioned above, this was triaged.
Hello!
Thank you for the amazing packages breakaway and DivNet.
I've been exploring the vignettes and functions in these packages and came across the function
phylodivnet
. I didn't see it in any of the vignettes but it appears to be for unifrac distance calculation for evaluating beta diversity? Can an example withphylodivnet
be added to the beta diversity vignette?I tried running
phylodivnet
with a phyloseq object that includes a phylogenetic tree and a diagonal design matrix like I would when runningdivnet
for beta diversity estimation and received the error... Error in dimnames(x) <- dn : length of 'dimnames' [2] not equal to array extentI'm running this on a dataset that I collapsed to the phylum level. Could this error be related to that? That the number of taxa doesn't reflect the tips of the phylogenetic tree?
Thanks!