SKILL.md
Version Compatibility
Reference examples tested with: DESeq2 1.42+, ggplot2 3.5+, phyloseq 1.46+, scanpy 1.10+
Before using code patterns, verify installed versions match. If versions differ:
- R:
packageVersion('<pkg>')then?function_nameto verify parameters
If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.
Differential Abundance Testing
"Find which taxa differ between my groups" → Identify differentially abundant taxa between experimental conditions using compositionally-aware methods that account for the relative nature of microbiome data.
- R:
ALDEx2::aldex()for CLR-transformed Welch's t-test - R:
ANCOMBC::ancombc2()for bias-corrected log-linear models - R:
Maaslin2::Maaslin2()for multivariable association
The Compositionality Problem
Microbiome data is compositional - abundances are relative, not absolute. Standard tests (t-test, DESeq2) can give false positives.
ALDEx2 (Recommended)
Goal: Identify differentially abundant taxa between groups using a compositionally-aware statistical framework.
Approach: Apply CLR transformation with Monte Carlo sampling on the OTU table, run Welch's t-test per taxon, and filter by FDR-corrected p-value and effect size.
library(ALDEx2)
library(phyloseq)
ps <- readRDS('phyloseq_object.rds')
otu <- as.data.frame(otu_table(ps))
if (!taxa_are_rows(ps)) otu <- t(otu)
# Define groups
groups <- sample_data(ps)$Group
# Run ALDEx2 (CLR transformation + Welch's t-test)
aldex_results <- aldex(otu, groups, mc.samples = 128, test = 'welch',
effect = TRUE, include.sample.summary = FALSE)
# Filter significant
sig_aldex <- aldex_results[aldex_results$we.eBH < 0.05 & abs(aldex_results$effect) > 1, ]
# Volcano-like plot
aldex.plot(aldex_results, type = 'MW', test = 'welch')