Reaching the End-Game for GWAS: Machine Learning Approaches for the Prioritization of Complex Disease Loci

Nicholls, Hannah; John, Christopher R.; Watson, David S.; Munroe, Patricia B.; Barnes, Michael R.; Cabrera, Claudia

doi:10.3389/fgene.2020.00350

Cited by 113 publications

(73 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given the complexity underlying the genetics of quantitative traits, it is probably not realistic to assume that any one method can retain its statistical power for different genetic architectures [ 17 , 35 , 36 ]. Single-SNP based models are still popular [ 37 , 38 , 39 , 40 , 41 ] while the RF based methods are gaining importance [ 42 ]. However, an increasing number of scientists are recommending the integration of different association methods in order to improve QTL identification and interpretation [ 43 , 44 ].…”

Section: Introductionmentioning

confidence: 99%

Combining Random Forests and a Signal Detection Method Leads to the Robust Detection of Genotype-Phenotype Associations

Ramzan

Gültaş

Bertram

et al. 2020

Genes

View full text Add to dashboard Cite

Genome wide association studies (GWAS) are a well established methodology to identify genomic variants and genes that are responsible for traits of interest in all branches of the life sciences. Despite the long time this methodology has had to mature the reliable detection of genotype–phenotype associations is still a challenge for many quantitative traits mainly because of the large number of genomic loci with weak individual effects on the trait under investigation. Thus, it can be hypothesized that many genomic variants that have a small, however real, effect remain unnoticed in many GWAS approaches. Here, we propose a two-step procedure to address this problem. In a first step, cubic splines are fitted to the test statistic values and genomic regions with spline-peaks that are higher than expected by chance are considered as quantitative trait loci (QTL). Then the SNPs in these QTLs are prioritized with respect to the strength of their association with the phenotype using a Random Forests approach. As a case study, we apply our procedure to real data sets and find trustworthy numbers of, partially novel, genomic variants and genes involved in various egg quality traits.

show abstract

Section: Introductionmentioning

confidence: 99%

Combining Random Forests and a Signal Detection Method Leads to the Robust Detection of Genotype-Phenotype Associations

Ramzan

Gültaş

Bertram

et al. 2020

Genes

View full text Add to dashboard Cite

show abstract

“…As far as the machine learning models are concerned, the best performing machine learning models which were utilized for comparison were different. However, the low reproducibility could also derive from a more general challenge of gene prioritization studies, which concerns the difficulty of identifying the truly causal genes out of a usually large pool of novel predicted genes that pass a chosen significance threshold [ 157 ]. The difficulty of reproducibility is also emphasized in Bean et al where 5 known ALS-linked gene lists mined from different databases, and one which was manually curated, were used, with each yielding very different results even with the same model, highlighting that the methodology and results of each study should be compared with caution [ 105 ].…”

Section: Discussionmentioning

confidence: 99%

What Can Machine Learning Approaches in Genomics Tell Us about the Molecular Basis of Amyotrophic Lateral Sclerosis?

Vasilopoulou

Morris

Giannakopoulos

et al. 2020

JPM

View full text Add to dashboard Cite

Amyotrophic Lateral Sclerosis (ALS) is the most common late-onset motor neuron disorder, but our current knowledge of the molecular mechanisms and pathways underlying this disease remain elusive. This review (1) systematically identifies machine learning studies aimed at the understanding of the genetic architecture of ALS, (2) outlines the main challenges faced and compares the different approaches that have been used to confront them, and (3) compares the experimental designs and results produced by those approaches and describes their reproducibility in terms of biological results and the performances of the machine learning models. The majority of the collected studies incorporated prior knowledge of ALS into their feature selection approaches, and trained their machine learning models using genomic data combined with other types of mined knowledge including functional associations, protein-protein interactions, disease/tissue-specific information, epigenetic data, and known ALS phenotype-genotype associations. The importance of incorporating gene-gene interactions and cis-regulatory elements into the experimental design of future ALS machine learning studies is highlighted. Lastly, it is suggested that future advances in the genomic and machine learning fields will bring about a better understanding of ALS genetic architecture, and enable improved personalized approaches to this and other devastating and complex diseases.

show abstract

“…Algorithms to identify footprints of selective sweeps in natural populations and genotype–phenotype associations are becoming available for the community ( Table 2 ). These analytical tools are expected to significantly improve the predictability of the causative mutation(s) through post hoc analysis especially in complex traits ( Ramstein et al, 2019 ; Nicholls et al, 2020 ). Other applications of ML algorithms can help to accelerate the breeding process through the implementation of deep learning methods in phenotyping, genomic selection, prediction of functionality, and so forth (reviewed in Wang et al, 2019 ).…”

Section: Future Directionsmentioning

confidence: 99%

Tapping Diversity From the Wild: From Sampling to Implementation

Hübner

Kantar

2021

Front. Plant Sci.

View full text Add to dashboard Cite

The diversity observed among crop wild relatives (CWRs) and their ability to flourish in unfavorable and harsh environments have drawn the attention of plant scientists and breeders for many decades. However, it is also recognized that the benefit gained from using CWRs in breeding is a potential rose between thorns of detrimental genetic variation that is linked to the trait of interest. Despite the increased interest in CWRs, little attention was given so far to the statistical, analytical, and technical considerations that should guide the sampling design, the germplasm characterization, and later its implementation in breeding. Here, we review the entire process of sampling and identifying beneficial genetic variation in CWRs and the challenge of using it in breeding. The ability to detect beneficial genetic variation in CWRs is strongly affected by the sampling design which should be adjusted to the spatial and temporal variation of the target species, the trait of interest, and the analytical approach used. Moreover, linkage disequilibrium is a key factor that constrains the resolution of searching for beneficial alleles along the genome, and later, the ability to deplete linked deleterious genetic variation as a consequence of genetic drag. We also discuss how technological advances in genomics, phenomics, biotechnology, and data science can improve the ability to identify beneficial genetic variation in CWRs and to exploit it in strive for higher-yielding and sustainable crops.

show abstract

Reaching the End-Game for GWAS: Machine Learning Approaches for the Prioritization of Complex Disease Loci

Cited by 113 publications

References 69 publications

Combining Random Forests and a Signal Detection Method Leads to the Robust Detection of Genotype-Phenotype Associations

Combining Random Forests and a Signal Detection Method Leads to the Robust Detection of Genotype-Phenotype Associations

What Can Machine Learning Approaches in Genomics Tell Us about the Molecular Basis of Amyotrophic Lateral Sclerosis?

Tapping Diversity From the Wild: From Sampling to Implementation

Contact Info

Product

Resources

About