Inference of genetic networks using random forests: Assigning different weights for gene expression data

Abstract

In using gene expression levels for genetic network inference, we believe that two measurements that are similar to each other are less informative than two measurements that differ from each other. Given, for example, that gene expression levels measured at two adjacent time points in a time-series experiment are often similar to each other, we assume that each measurement in the time-series experiment will be less informative than each measurement in a steady-state experiment. Based on this idea, we propose a new inference method that relies heavily on informative gene expression data. Through numerical experiments, we prove that the quality of an inferred genetic network is slightly improved by heavily weighting informative gene expression data. In this study, we develop a new method by modifying the existing random-forest-based inference method to take advantage of its ability to analyze both time-series and static gene expression data. The idea we propose can be similarly applied to many of the other existing inference methods, as well.

Inference of Genetic Networks From Time-Series and Static Gene Expression Data: Combining a Random-Forest-Based Inference Method With Feature Selection Methods

“…As mentioned previously, this study combines the randomforest-based inference method with a series of feature selection methods. While any random-forest-based method can serve this purpose, in this study we apply an inference method (Kimura et al, 2019) that is capable of analyzing both time-series and static gene expression data. This section briefly describes the inference method.…”

“…The inference method (Kimura et al, 2019) divides an inference problem of a genetic network consisting of N genes into N subproblems, each of which corresponds to each gene. By solving the n-th subproblem, the method obtains a reasonable approximation of the function F n and a reasonable value for the parameter β n .…”

“…w T k and w S k are weight parameters for the k-th measurements in the timeseries and steady-state experiments, respectively. Kimura et al (2019) showed that the performance of the random-forest-based inference method improves by discounting the weight values of the measurements that were obtained under states similar to each other.F n ( • ; β n ) is an approximation of the function F n trained under the given β n . The inference method (Kimura et al, 2019) obtains an approximation of the function F n using a random forest (Breiman, 2001).…”

“…A number of genetic network inference methods have been proposed (Larrañaga et al, 2006;Meyer et al, 2008;Chou and Voit, 2009;Hecker et al, 2009;de Matos Simoes and Emmert-Streib, 2012;Emmert-Streib et al, 2012;Glass et al, 2013). Among them, random-forest-based methods show promise for their excellent performance (Huynh-Thu et al, 2010;Maduranga et al, 2013;Petralia et al, 2015;Huynh-Thu and Geurts, 2018;Kimura et al, 2019). Some of these inference methods also have a useful ability to analyze both time-series and static gene expression data (Petralia et al, 2015;Huynh-Thu and Geurts, 2018;Kimura et al, 2019).…”

“…Among them, random-forest-based methods show promise for their excellent performance (Huynh-Thu et al, 2010;Maduranga et al, 2013;Petralia et al, 2015;Huynh-Thu and Geurts, 2018;Kimura et al, 2019). Some of these inference methods also have a useful ability to analyze both time-series and static gene expression data (Petralia et al, 2015;Huynh-Thu and Geurts, 2018;Kimura et al, 2019). The time-series data are a series of sets of gene expression levels measured at successive time points after a stimulation.…”