Data mining in the Life Sciences with Random Forest: a walk in the park or lost in the jungle?

Abstract

In the Life Sciences ‘omics’ data is increasingly generated by different high-throughput technologies. Often only the integration of these data allows uncovering biological insights that can be experimentally validated or mechanistically modelled, i.e. sophisticated computational approaches are required to extract the complex non-linear trends present in omics data. Classification techniques allow training a model based on variables (e.g. SNPs in genetic association studies) to separate different classes (e.g. healthy subjects versus patients). Random Forest (RF) is a versatile classification algorithm suited for the analysis of these large data sets. In the Life Sciences, RF is popular because RF classification models have a high-prediction accuracy and provide information on importance of variables for classification. For omics data, variables or conditional relations between variables are typically important for a subset of samples of the same class. For example: within a class of cancer patients certain SNP combinations may be important for a subset of patients that have a specific subtype of cancer, but not important for a different subset of patients. These conditional relationships can in principle be uncovered from the data with RF as these are implicitly taken into account by the algorithm during the creation of the classification model. This review details some of the to the best of our knowledge rarely or never used RF properties that allow maximizing the biological insights that can be extracted from complex omics data sets using RF.

T lymphocytes control microbial composition by regulating the abundance of Vibrio in the zebrafish gut

“…31 In our analysis, we considered an OTU to be highly predictive if its importance score was at least 0.001. 32 ( Figure S2). …”

Machine Learning–Based Differential Network Analysis: A Study of Stress-Responsive Transcriptomes in Arabidopsis

“…The RF algorithm builds thousands of decision trees with bootstrapped positive and negative samples and randomly selected characteristics in the input feature matrix (Breiman, 2001). This strategy can robustly reduce the influence from noise (the mislabeled positive or negative samples) and outliers (extremely high or low feature values) (Touw et al, 2013). The feature matrix submitted to the RF classifier included 12 characteristics of absolute expression values of a gene at six time points in control and stress situations, 12 characteristics of within-condition expression variations of a gene measured as z-scores at six time points in control and stress situations, six characteristics of between-condition expression changes of a gene measured as fold changes at six time points involving stress versus the control, and two characteristics of the coefficient of variation (CV) in stress and control situations (see Methods).…”

Ontology-Based Classification of Building Types Detected from Airborne Laser Scanning Data

“…To address the above-mentioned problems, we used the RF classifier [55]. RF is a non-parametric ensemble learning classifier [55], successfully implemented in different application domains, including remote sensing [56][57][58] and data mining in life sciences [59]. For a detailed evaluation of the effectiveness of the RF classifier in the remote sensing domain, the readers might refer to [60].…”