Local biplots for multidimensional scaling, with application to the microbiome

We present local biplots, an extension of the classic principal component biplot to multidimensional scaling. Noticing that principal component biplots have an interpretation as the Jacobian of a map from data space to the principal subspace, we define local biplots as the Jacobian of the analogous map for multidimensional scaling. In the process, we show a close relationship between our local biplot axes, generalized Euclidean distances, and generalized principal components. In simulations and real data we show how local biplots can shed light on what variables or combinations of variables are important for the low-dimensional embedding provided by multidimensional scaling. They give particular insight into a class of phylogenetically-informed distances commonly used in the analysis of microbiome data, showing that different variants of these distances can be interpreted as implicitly smoothing the data along the phylogenetic tree and that the extent of this smoothing is variable.

Automated summary

This paper presents the concept and application of local biplots. By analyzing the analogous mapping in multidimensional scaling, local biplots can reveal important variables or combinations of variables in the data and help explain the implicit data smoothing level in different distances.