No pdf for feature associations when only 1 species is significant

[apalleja]

Hi SIAMCAT developers,

I like very much your package! However when I run: plot.title=paste0("assoc_", analysis, ".pdf") check.associations( siamcat, sort.by = 'fc', fn.plot = plot.title, alpha = 0.1, mult.corr = "fdr", detect.lim = 10 ^ -6, max.show = 50, plot.type = "quantile.box", panels = c("fc", "prevalence", "auroc"), verbose = 2 )

I get this error and don't get any pdf for the significant species. It does not happen if I set the alpha much higher because then I have 3 species significant.

• starting check.associations
• starting analyse.binary.marker +++ calculating effect size for each feature. |=======================================================================================================================================| 100%+++ found 1 significant associations at a significance level < 0.1
• finished analyse.binary.marker in 0.356 s Error: Argument 'x' is of class 'numeric', but should be a matrix. The use of a 'numeric' is not supported, the correctness of the result is not guaranteed. Please update your code accordingly. In addition: Warning message: In analyse.binary.marker(feat = feat, label = label, detect.lim = detect.lim, : Less than 5 associations found. Consider changing your alpha value

Please let me know if there is a quick fix for this ;P

Albert

[jakob-wirbel]

Hello Albert,

the quickest fix would be to up your alpha a bit, because it seems that there are fewer than 5 significant associations at your chosen alpha level. You could try alpha=0.2 in your function call (or any higher alpha) and check the results.

But indeed, there seems to be a bug which we will try to fix soon :-)

Cheers, Jakob

[jakob-wirbel]

The association plot should work already when you chose as plot.type one of these two:

• box
• bean

The quantile plots should be fixed with this commit now, i hope 😄