Random forest fishing: a novel approach to identifying organic group of risk factors in genome-wide association studies

Yang, Wei; Gu, C. Charles

doi:10.1038/ejhg.2013.109

Cited by 7 publications

(6 citation statements)

References 30 publications

(23 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning methods such as multi‐dimensional reduction (MDR), support vector machines (SVM), neural networks (NN) and random forest (RF) are capable of dealing with this dimensionality problem in a flexible manner and can effectively select important variables from irrelevant ones (Goldstein et al . ; Gonzalez‐Recio & Forni ; González‐Recio, Rosa & Gianola ; Yang & Charles Gu ). In particular, RF analysis has been particularly useful in pathway analysis because interactions are implicitly modelled (De Lobel et al .…”

Section: Narrowing Down Adaptations At the Dna Levelmentioning

confidence: 99%

Towards the identification of the loci of adaptive evolution

2015

View full text Add to dashboard Cite

1. Establishing the genetic and molecular basis underlying adaptive traits is one of the major goals of evolutionary geneticists in order to understand the connection between genotype and phenotype and elucidate the mechanisms of evolutionary change. Despite considerable effort to address this question, there remain relatively few systems in which the genes shaping adaptations have been identified.2. Here, we review the experimental tools that have been applied to document the molecular basis underlying evolution in several natural systems, in order to highlight their benefits, limitations and suitability. In most cases, a combination of DNA, RNA and functional methodologies with field experiments will be needed to uncover the genes and mechanisms shaping adaptation in nature.

show abstract

Section: Narrowing Down Adaptations At the Dna Levelmentioning

confidence: 99%

Towards the identification of the loci of adaptive evolution

2015

View full text Add to dashboard Cite

show abstract

“…For instance, permuted random forest (pRF) identifies interacting SNP pairs by systematically permuting interactions between a pair of SNPs and determining which SNP pairs cause the greatest reduction in prediction power (Li et al 2016 ). Random forest fishing (RFF) is an iterative approach that has been shown to identify important variants even when no main effects are present on the variants (Yang and Charles Gu 2014 ).…”

Section: Methodsmentioning

confidence: 99%

What makes a good prediction? Feature importance and beginning to open the black box of machine learning in genetics

et al. 2021

View full text Add to dashboard Cite

Genetic data have become increasingly complex within the past decade, leading researchers to pursue increasingly complex questions, such as those involving epistatic interactions and protein prediction. Traditional methods are ill-suited to answer these questions, but machine learning (ML) techniques offer an alternative solution. ML algorithms are commonly used in genetics to predict or classify subjects, but some methods evaluate which features (variables) are responsible for creating a good prediction; this is called feature importance. This is critical in genetics, as researchers are often interested in which features (e.g., SNP genotype or environmental exposure) are responsible for a good prediction. This allows for the deeper analysis beyond simple prediction, including the determination of risk factors associated with a given phenotype. Feature importance further permits the researcher to peer inside the black box of many ML algorithms to see how they work and which features are critical in informing a good prediction. This review focuses on ML methods that provide feature importance metrics for the analysis of genetic data. Five major categories of ML algorithms: k nearest neighbors, artificial neural networks, deep learning, support vector machines, and random forests are described. The review ends with a discussion of how to choose the best machine for a data set. This review will be particularly useful for genetic researchers looking to use ML methods to answer questions beyond basic prediction and classification.

show abstract

“…People have successfully developed tree-based methods to infer gene regulatory networks [ 19 ]. Based on RF, Random Forest Fishing (RFF) has been designed to effectively identify risk factors when considering both marginal effects and interactions using GWAS data [ 20 ]. A software package named Random Jungle (RJ) was also designed typically for large-scale association studies which is a fast-implementation of RF and Cordell has applied real data analysis using this package to identify gene-gene interactions [ 21 , 22 ].…”

Section: Introductionmentioning

confidence: 99%

Detecting gene-gene interactions using a permutation-based random forest method

Malley

Andrew

et al. 2016

BioData Mining

View full text Add to dashboard Cite

BackgroundIdentifying gene-gene interactions is essential to understand disease susceptibility and to detect genetic architectures underlying complex diseases. Here, we aimed at developing a permutation-based methodology relying on a machine learning method, random forest (RF), to detect gene-gene interactions. Our approach called permuted random forest (pRF) which identified the top interacting single nucleotide polymorphism (SNP) pairs by estimating how much the power of a random forest classification model is influenced by removing pairwise interactions.ResultsWe systematically tested our approach on a simulation study with datasets possessing various genetic constraints including heritability, number of SNPs, sample size, etc. Our methodology showed high success rates for detecting the interaction SNP pair. We also applied our approach to two bladder cancer datasets, which showed consistent results with well-studied methodologies, such as multifactor dimensionality reduction (MDR) and statistical epistasis network (SEN). Furthermore, we built permuted random forest networks (PRFN), in which we used nodes to represent SNPs and edges to indicate interactions.ConclusionsWe successfully developed a scale-invariant methodology to detect pure gene-gene interactions based on permutation strategies and the machine learning method random forest. This methodology showed great potential to be used for detecting gene-gene interactions to study underlying genetic architectures in a scale-free way, which could be benefit to uncover the complex disease mechanisms.

show abstract

Random forest fishing: a novel approach to identifying organic group of risk factors in genome-wide association studies

Cited by 7 publications

References 30 publications

Towards the identification of the loci of adaptive evolution

Towards the identification of the loci of adaptive evolution

What makes a good prediction? Feature importance and beginning to open the black box of machine learning in genetics

Detecting gene-gene interactions using a permutation-based random forest method

Contact Info

Product

Resources

About