IFAA - Robust Inference for Absolute Abundance in Microbiome Analysis
This package offers a robust approach to make inference on
the association of covariates with the absolute abundance (AA)
of microbiome in an ecosystem. It can be also directly applied
to relative abundance (RA) data to make inference on AA because
the ratio of two RA is equal to the ratio of their AA. This
algorithm can estimate and test the associations of interest
while adjusting for potential confounders. The estimates of
this method have easy interpretation like a typical regression
analysis. High-dimensional covariates are handled with
regularization and it is implemented by parallel computing.
False discovery rate is automatically controlled by this
approach. Zeros do not need to be imputed by a positive value
for the analysis. The IFAA package also offers the 'MZILN'
function for estimating and testing associations of abundance
ratios with covariates.