SuRF: A new method for sparse variable selection, with application in microbiome data analysis

Abstract

In this article, we present a new variable selection method for regression and classification purposes, particularly for microbiome analysis. Our method, called subsampling ranking forward selection (SuRF), is based on LASSO penalized regression, subsampling and forward‐selection methods. SuRF offers major advantages over existing variable selection methods in terms of both sparsity of selected models and model inference. We provide an R package that can implement our method for generalized linear models. We apply our method to classification problems from microbiome data, using a novel agglomeration approach to deal with the special tree‐like correlation structure of the variables. Existing methods arbitrarily choose a taxonomic level a priori before performing the analysis, whereas by combining SuRF with these aggregated variables, we are able to identify the key biomarkers at the appropriate taxonomic level, as suggested by the data. We present simulations in multiple sparse settings to demonstrate that our approach performs better than several other popularly used existing approaches in recovering the true variables. We apply SuRF to two microbiome datasets: one about prediction of pouchitis and another for identifying samples from two healthy individuals. We find that SuRF can provide a better or comparable prediction with other methods while controlling the false positive rate of variable selection.

NFnetFu: A novel workflow for microbiome data fusion

“…LEfSe focuses on ranking based on how efficiently features discriminate with respect to the classes whereas NFnetFu aims to rank features with respect to their association with the outcome variable. SuRF [ 62 ] is more advantageous in comparison to the existing methods for variable selection in terms of dealing with model inference and sparsity of selected models. Since its variable selection is based on a LASSO based approach, similar to the approach NFnetFU adopts, the results across the two tools are more comparable.…”

“…This method includes subsampling and forward-selection methods which primarily focus on microbiome analysis. The SuRF [ 62 ] framework consists of mainly two steps. Firstly, an ordered list of predictors, using the LASSO variable selection method, is generated over subsampled observations.…”

Variable Selection with Scalable Bootstrapping in Generalized Linear Model for Massive Data

“…Meinshausen and Bühlmann (2010) introduced a stability selection framework for high-dimensional data, based on the subsampling algorithm. Liu et al (2021) applied the subsampling strategy to lasso penalized regression to analysis microbiome data. A closely related method is the m out of n bootstrap proposed by (Bickel et al, 2012), which randomly samples m < n observations from the original dataset.…”