Open 411an13 opened 6 months ago
Dear @411an13, I am a user just like you. And just like you I encountered the same issue. I also think that the issue belongs to the interface with the TSE object. Maybe it's a version incompatibility issue? I didn't investigate.
I suggest you to create a phyloseq object from the TSE with the utility function from mia package: phyloseq_object_name <- mia::makePhyloseqFromTreeSummarizedExperiment(tse_object_name)
Best,
Context
I'm trying to run ANCOM-BC2 for the first time and I keep getting the same error message:
Error in is.infinite(o1) : default method not implemented for type 'list'
. I'm wondering if this has been reported before (to my knowledge, it hasn't been) or if there's a simple syntax error in my implementation. I've checked the documentation and ANCOM-BC2 tutorial. Briefly, I'm interested in analyzing the effects of two categorical variables:env
(2 levels) andstrain
(3 levels). For each combination of these variables, I have three replicates, so total N = 18 (2x3x3). Initial visualizations suggest strong segregation by treatment; now I aim to quantify the effects.Issue
I've put my data/metadata into a TreeSummarizedExperiment format, and I'm not interested in doing a trend test but I've also run the following code with the trend test lines included, which still resulted in the same error message. Is there an apparent error in how I'm using the function? If not, I suspect it could be an issue related to the TSE itself. I just can't tell what the error message is referring to.
Attempted fixes
Fix: Uncomment the trend_control parameter
Result: Same error, regardless of whether
trend
was set to TRUE or FALSEFix: Set
trend
to TRUE but keeptrend_control
commented out to see if the error is triggered before/afterResult:
Error: Please specify the contrast matrices for the trend test.
My interpretation is that the error occurs after arguments are read in.Fix: Setting
global
,pairwise
, anddunnet
to FALSE individually and all togetherResult: Same error
Session info