TSLRF: Two-Stage Algorithm Based on Least Angle Regression and Random Forest in genome-wide association studies

Abstract

One of the most important tasks in genome-wide association analysis (GWAS) is the detection of single-nucleotide polymorphisms (SNPs) which are related to target traits. With the development of sequencing technology, traditional statistical methods are difficult to analyze the corresponding high-dimensional massive data or SNPs. Recently, machine learning methods have become more popular in high-dimensional genetic data analysis for their fast computation speed. However, most of machine learning methods have several drawbacks, such as poor generalization ability, over-fitting, unsatisfactory classification and low detection accuracy. This study proposed a two-stage algorithm based on least angle regression and random forest (TSLRF), which firstly considered the control of population structure and polygenic effects, then selected the SNPs that were potentially related to target traits by using least angle regression (LARS), furtherly analyzed this variable subset using random forest (RF) to detect quantitative trait nucleotides (QTNs) associated with target traits. The new method has more powerful detection in simulation experiments and real data analyses. The results of simulation experiments showed that, compared with the existing approaches, the new method effectively improved the detection ability of QTNs and model fitting degree, and required less calculation time. In addition, the new method significantly distinguished QTNs and other SNPs. Subsequently, the new method was applied to analyze five flowering-related traits in Arabidopsis. The results showed that, the distinction between QTNs and unrelated SNPs was more significant than the other methods. The new method detected 60 genes confirmed to be related to the target trait, which was significantly higher than the other methods, and simultaneously detected multiple gene clusters associated with the target trait.

Proposed Two-Steps Procedure of Classification High Dimensional Data with Regularized Logistic Regression

“…One of the study topics in statistical applications is the reduction of dimensionality since this is often the most effective method for working with large datasets of high complexity [7,8]. Over the last decade, the disciplines of machine learning and statistical methods have seen the introduction of many two-stage procedures, such as ranking variable selection (RVS) and classification algorithms, aimed at effectively tackling these challenges [5,9,10]. The application of RVS techniques in classification methodologies consistently improves performance by reducing data dimensionality through the exclusion of irrelevant genes (variables).…”

Variable Selection in High Dimensional Data with Interactions

“…Thus, the proposed methods apply to the multi-pollutant framework, and the method used should be based goal of the study, be it a prediction, effect estimation, or screening for significant predictors and their interactions. Jiali S. [12] proposed a two-stage algorithm based on Least Angle Regression (LARS) and Random Forests. They concluded that the proposed method significantly improved the model fitting and variable selection, requiring less calculation time.…”

Intersectionality in HR research: challenges and opportunities

“…This approach of detecting key intersecting categories in a training sample and inputted into a subsequent analysis sample has been used in genome researchfor which extremely high-dimensional data are the normfor detecting the intersecting combinations of genes that predict diseases. Examples include Khan et al (2020), Sun et al (2019) and Wright et al (2016). Even if key predictor interactions could be detected by machine learning, it does not seem possible to determine whether any statistically unimportant effects are due to true absences of effects or an underpowered analysis due to few observations in the intersecting categories.…”