PANDA: Prioritization of autism‐genes using network‐based deep‐learning approach

Chen, Yuanzhu; Hu, Ting

doi:10.1002/gepi.22282

Cited by 18 publications

(16 citation statements)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…PANDA (Y. Zhang et al, 2020) is a graph neural network type classifier. They built an unweighted and undirected human molecular interaction network from experimentally documented physical protein interactions using data from a previously established protein-protein interaction network (Menche et al, 2015), and BioGRID (Oughtred et al, 2019).…”

Section: Gba ML Methodsmentioning

confidence: 99%

“…Elucidating the genetic architecture of complex human disorders and diseases is currently a major challenge in medical research. Identifying genes involved in disease is often a time consuming and expensive process, so many researchers have been attracted to the idea of using predictions generated by machine learning (ML) algorithms (Krishnan et al, 2016;Lee et al, 2011;Moreau & Tranchevent, 2012;Y. Zhang et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

“…Recently there has been interest in a specific use case for GBA-based ML: predicting genes responsible for genetic risk of autism spectrum disorder (ASD). Multiple GBA-based ML studies have been produced with claims of providing greater insight into the genetic etiology of ASD (Brueggeman et al, 2020;Duda et al, 2018;Krishnan et al, 2016;Lin et al, 2018;Y. Zhang et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Can machine learning aid in identifying disease genes? The case of autism spectrum disorder

Gunning

Pavlidis

2020

Preprint

View full text Add to dashboard Cite

Discovering genes involved in complex human genetic disorders is a major challenge. Many have suggested that machine learning (ML) algorithms using gene networks can be used to supplement traditional genetic association-based approaches to predict or prioritize disease genes. However, questions have been raised about the utility of ML methods for this type of task due to biases within the data, and poor real-world performance. Using autism spectrum disorder (ASD) as a test case, we sought to investigate the question: Can machine learning aid in the discovery of disease genes? We collected thirteen published ASD gene prioritization studies and evaluated their performance using known and novel high-confidence ASD genes. We also investigated their biases towards generic gene annotations, like number of association publications. We found that ML methods which do not incorporate genetics information have limited utility for prioritization of ASD risk genes. These studies perform at a comparable level to generic measures of likelihood for the involvement of genes in any condition, and do not out-perform genetic association studies. Future efforts to discover disease genes should be focused on developing and validating statistical models for genetic association, specifically for association between rare variants and disease, rather than developing complex machine learning methods using complex heterogeneous biological data with unknown reliability.

show abstract

Section: Gba ML Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Can machine learning aid in identifying disease genes? The case of autism spectrum disorder

Gunning

Pavlidis

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Several authors have investigated possible methods of producing network-based prioritization of ASD genes [16]. In particular, machine learning algorithms have been employed in order to delineate the ASD architecture [18][19][20]. In addition, cluster analysis has seen increasing applications in biomedicine, to further the understanding of ASD [21].…”

Section: Introductionmentioning

confidence: 99%

Clustering Analysis Supports the Detection of Biological Processes Related to Autism Spectrum Disorder

Gialloreti

Enea²,

Micco

et al. 2020

Genes

View full text Add to dashboard Cite

Genome sequencing has identified a large number of putative autism spectrum disorder (ASD) risk genes, revealing possible disrupted biological pathways; however, the genetic and environmental underpinnings of ASD remain mostly unanswered. The presented methodology aimed to identify genetically related clusters of ASD individuals. By using the VariCarta dataset, which contains data retrieved from 13,069 people with ASD, we compared patients pairwise to build “patient similarity matrices”. Hierarchical-agglomerative-clustering and heatmapping were performed, followed by enrichment analysis (EA). We analyzed whole-genome sequencing retrieved from 2062 individuals, and isolated 11,609 genetic variants shared by at least two people. The analysis yielded three clusters, composed, respectively, by 574 (27.8%), 507 (24.6%), and 650 (31.5%) individuals. Overall, 4187 variants (36.1%) were common to the three clusters. The EA revealed that the biological processes related to the shared genetic variants were mainly involved in neuron projection guidance and morphogenesis, cell junctions, synapse assembly, and in observational, imitative, and vocal learning. The study highlighted genetic networks, which were more frequent in a sample of people with ASD, compared to the overall population. We suggest that itemizing not only single variants, but also gene networks, might support ASD etiopathology research. Future work on larger databases will have to ascertain the reproducibility of this methodology.

show abstract

“…Several authors have investigated possible methods to produce network-based prioritization of ASD genes [15]. In particular, machine learning algorithms have been employed in order to delineate the ASD architecture [17] [18] [19]. Also cluster analysis has seen increasing applications in biomedicine to further the understanding of ASD [20].…”

Section: Introductionmentioning

confidence: 99%

A novel data-driven approach reveals gene networks and biological processes underlying autism

Gialloreti¹,

Enea²,

Micco³

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: Developments in gene-hunting techniques identified several ASD associated genes. The considerable significance of cluster analysis associated with gene network studies has led to reveal many disrupted key pathways in ASD, even if its genetic underpinnings remain a challenging task. This study aims to determine, through a novel data-driven approach, how networks of mutated genes impact biological processes underlying autism. Methods: We analyzed the VariCarta dataset, which presents more than 200,000 genomic variant events collected from 13,069 people with ASD. Firstly, we created a whole-genome and an exome sequencing subset. Then, for each subset we compared pairwise patients of each group to build “patient similarity matrices”. Hierarchical-agglomerative-clustering and heatmap were performed to identify clusters of patients with common occurrences of gene networks within these matrices. The subsequent enrichment analysis (EA) highlighted biological processes that might be impacted by the mutated genes of each subgroup. Results: Considering the whole-genome matrix, we identified three main genetic clusters of ASD patients, each one characterized by a network of shared genetic variants. We isolated 11,609 genetic variants shared by at least two subjects in each cluster; 4,187 of these variants (36.1%) were common to the three clusters. Only 331 patients (2.5%) shared none or very few mutated genes with anyone else. The EA highlighted common or cluster-specific biological processes related to the variants. Most of the common abnormal processes were involved in neuron projections guidance and morphogenesis, cell junctions and synapse assembly. Exome sequencing alone was not effectual in identifying ASD subgroups. Limitations: Caution is warranted when interpreting our results, as we did not compare them with a control group and did not verify if the identified subgroups where actually associated with different phenotypes. Future work will have to ascertain the strength and reproducibility of these results. Conclusions: Itemizing not just single mutated genes, but also gene networks and specific biological processes that characterize different ASD subpopulations might allow to better understand which networks of genetic variants play a major role in the etiopathology of ASD. The proposed methodology may represent a novel approach to help disentangle ASD complexity and an instrument to boost more focused genotype-phenotype studies.

show abstract

PANDA: Prioritization of autism‐genes using network‐based deep‐learning approach

Cited by 18 publications

References 65 publications

Can machine learning aid in identifying disease genes? The case of autism spectrum disorder

Can machine learning aid in identifying disease genes? The case of autism spectrum disorder

Clustering Analysis Supports the Detection of Biological Processes Related to Autism Spectrum Disorder

A novel data-driven approach reveals gene networks and biological processes underlying autism

Contact Info

Product

Resources

About