Biology subjects

Beyer, A.

Publications and source records attributed to Beyer, A..

At least 37 records · Page 2Linked to original sources

bioRxiv2021

Optimizing Network Propagation for Multi-Omics Data Integration

Network propagation refers to a class of algorithms that integrate information from input data across connected nodes in a given network. These algorithms have wide applications in systems biology, protein function prediction, inferring condition-specifically altered sub-networks, and prioritizing disease genes. Despite the popularity of network propagation, there is a lack of comparative analyses of different algorithms on real data and little guidance on how to select and parameterize the various algorithms. Here, we address this problem by analyzing different combinations of network normalization and propagation methods and by demonstrating schemes for the identification of optimal parameter settings on real proteome and transcriptome data. Our work highlights the risk of a topology bias caused by the incorrect use of network normalization approaches. Capitalizing on the fact that network propagation is a regularization approach, we show that minimizing the bias-variance tradeoff can be utilized for selecting optimal parameters. The application to real multi-omics data demonstrated that optimal parameters could also be obtained by either maximizing the agreement between different omics layers (e.g. proteome and transcriptome) or by maximizing the consistency between biological replicates. Furthermore, we exemplified the utility and robustness of network propagation on multi-omics datasets for identifying ageing-associated genes in brain and liver tissues of rats and for elucidating molecular mechanisms underlying prostate cancer progression. Overall, this work compares different network propagation approaches and it presents strategies for how to use network propagation algorithms to optimally address a specific research question at hand.

bioinformatics↗

bioRxiv2021

Biological mechanisms of aging predict age-related disease multimorbidities in patients

Genetic, environmental and pharmacological interventions into the aging process can confer resistance to a multiple age-related diseases in laboratory animals, including rhesus monkeys. These findings imply that mechanisms of aging might contribute to patterns of multimorbidity in humans, and hence could be targeted to prevent multiple conditions simultaneously. To address this question, we text mined 917,645 literature abstracts followed by manual curation, and found strong, non-random associations between age-related diseases and aging mechanisms, confirmed by gene set enrichment analysis of GWAS data. Integration of these associations with clinical data from 3.01 million patients showed that age-related diseases associated with each of five aging mechanisms were more likely than chance to be present together in patients. Genetic evidence revealed that innate and adaptive immunity, the intrinsic apoptotic signalling pathway and activity of the ERK1/2 pathway played a significant role across multiple aging mechanisms and multiple, diverse age-related diseases. Mechanisms of aging therefore contribute to multiple age-related diseases and to patterns of human age-related multimorbidity, and could potentially be targeted to prevent more than one age-related condition in the same patient.

bioinformatics↗

bioRxiv2020

Novel imprints in mouse blastocysts are predominantly DNA methylation independent

In mammals, chromatin marks at imprinted genes are asymmetrically inherited to control parentally-biased gene expression. This control is thought predominantly to involve parent-specific differentially methylated regions (DMR) in genomic DNA. However, neither parent-of-origin-specific transcription nor DMRs have been comprehensively mapped. We here address this by integrating transcriptomic and epigenomic approaches in mouse preimplantation embryos (blastocysts). Transcriptome-analysis identified 71 genes expressed with previously unknown parent-of-origin-specific expression in blastocysts (nBiX: novel blastocyst-imprinted expression). Uniparental expression of nBiX genes disappeared soon after implantation. Micro-whole-genome bisulfite sequencing (WGBS) of individual uniparental blastocysts detected 859 DMRs. Only 18% of nBiXs were associated with a DMR, whereas 60% were associated with parentally-biased H3K27me3. This suggests a major role for Polycomb-mediated imprinting in blastocysts. Five nBiX-clusters contained at least one known imprinted gene, and five novel clusters contained exclusively nBiX-genes. These data suggest a complex program of stage-specific imprinting involving different tiers of regulation.

genomics↗

bioRxiv2020

NMD is required for timely cell fate transitions by fine-tuning gene expression and controlling translation

Cell fate transitions depend on balanced rewiring of transcription and translation programmes to mediate ordered developmental progression. Here we identify a feedback loop between nonsense-mediated mRNA decay (NMD) and translation initiation. We show that NMD controls the translation initiation factor Eif4a2 and its premature termination codon encoding isoform (Eif4a2PTC). NMD deficiency leads to translation of a specific truncated Eif4a2 protein, which elicits increased translation rates and is causative for significant delays in mouse embryonic stem cell (ESC) differentiation. Our results show identical mRNA targets for Smg5, Smg6 and Smg7, but illustrate a clear hierarchy between KOs in amplitude of target deregulation and differentiation phenotype (Smg5 > Smg6 > Smg7). This hierarchy highlights heterodimer independent functions for Smg5 and Smg7. Together, our findings expose an intricate link between mRNA stability and translation initiation control, that must be maintained for normal dynamics of cell state transitions.Competing Interest StatementThe authors have declared no competing interest.View Full Text

cell biology↗

bioRxiv2020

Cis-regulatory evolution spotlights species differences in the adaptive potential of gene expression plasticity

Phenotypic plasticity is the variation in phenotype that a single genotype can produce in different environments and, as such, is an important component of individual fitness. However, whether the effect of new mutations, and hence evolution, depends on the direction of plasticity remains controversial. Here, we identify the cis-acting modifications that have reshaped gene expression in response to dehydration stress in three Arabidopsis species. Our study shows that the direction of effects of most cis-regulatory variants differentiating the response between A. thaliana and the sister species A. lyrata and A. halleri depends on the direction of pre-existing plasticity in gene expression. A comparison of the rate of cis-acting variant accumulation in each lineage indicates that the selective forces driving adaptive evolution in gene expression favors regulatory changes that magnify the stress response in A. lyrata. The evolutionary constraints measured on the amino-acid sequence of these genes support this interpretation. In contrast, regulatory changes that mitigate the plastic response to stress evolved more frequently in A. halleri. Our results demonstrate that pre-existing plasticity can be a stepping stone for adaptation, but its selective remodeling differs between lineages.

evolutionary biology↗

bioRxiv2020

Cooperative molecular networks drive a mammalian cell state transition

In the mammalian embryo, epiblast cells must exit their naive state and acquire formative pluripotency. This cell state transition is recapitulated by mouse embryonic stem cells (ESCs), which undergo pluripotency progression in defined conditions in vitro. However, our understanding of the molecular cascades and gene-networks involved in the exit from naive pluripotency remains fragmented. Here we employed a combination of genetic screens in haploid ESCs, CRISPR/Cas9 gene disruption, large-scale transcriptomics and computational systems-biology to delineate the regulatory circuits governing naive state exit. Transcriptome profiles for 73 knockout ESC lines predominantly manifest delays on the trajectory from naive to formative epiblast. We find that gene networks operative in ESCs are active during transition from pre- to post-implantation epiblast in utero. We identified 374 naive-associated genes tightly connected to epiblast state and largely conserved in human ESCs and primate embryos. Integrated analysis of mutant transcriptomes revealed funneling of multiple gene activities into discrete regulatory modules. Finally, we delineate how intersections with signaling pathways direct this pivotal mammalian cell state transition.

systems biology↗

bioRxiv2020

Proteogenomic heterogeneity of localized human prostate cancer progression

BackgroundTumor-specific genomic aberrations are routinely determined by high throughput genomic measurements. It remains unclear though, how complex genome alterations affect molecular networks through changing protein levels, and consequently biochemical states of tumor tissues. ResultsHere, we investigated the propagation of genomic effects along the axis of gene expression during prostate cancer progression. For that, we quantified genomic, transcriptomic and proteomic alterations based on 105 prostate samples, consisting of benign prostatic hyperplasia regions and malignant tumors, from 39 prostate cancer patients. Our analysis revealed convergent effects of distinct copy number alterations impacting on common downstream proteins, which are important for establishing the tumor phenotype. We devised a network-based approach that integrates perturbations across different molecular layers, which identified a sub-network consisting of nine genes whose joint activity positively correlated with increasingly aggressive tumor phenotypes and was predictive of recurrence-free survival. Further, our data revealed a wide spectrum of intra-patient network effects, ranging from similar to very distinct alterations on different molecular layers. ConclusionsThis study uncovered molecular networks with remarkably convergent alterations across tumor sites and patients, but it also exposed a diversity of network effects: we could not identify a single sub-network that was perturbed in all high-grade tumor regions.

systems biology↗

bioRxiv2020

Deciphering the Signaling Network Landscape of Breast Cancer Improves Drug Sensitivity Prediction

Although genetic and epigenetic abnormalities in breast cancer have been extensively studied, it remains difficult to identify those patients who will respond to particular therapies. This is due in part to our lack of understanding of how the variability of cellular signaling affects drug sensitivity. Here, we used mass cytometry to characterize the single-cell signaling landscapes of 62 breast cancer cell lines and five lines from healthy tissue. We quantified 34 markers in each cell line upon stimulation by the growth factor EGF in the presence or absence of five kinase inhibitors. These data - on more than 80 million single cells from 4,000 conditions - were used to fit mechanistic signaling network models that provide unprecedented insights into the biological principles of how cancer cells process information. Our dynamic single-cell-based models more accurately predicted drug sensitivity than static bulk measurements for drugs targeting the PI3K-MTOR signaling pathway. Finally, we identified genomic features associated with drug sensitivity by using signaling phenotypes as proxies, including a missense mutation in DDIT3 predictive of PI3K-inhibition sensitivity. This provides proof of principle that single-cell measurements and modeling could inform matching of patients with appropriate treatments in the future. One-linerSingle-cell proteomics coupled to perturbations improves accuracy of breast tumor drug sensitivity predictions and reveals mechanisms of sensitivity and resistance. HIGHLIGHTSO_LIMass cytometry study of signaling responses of 62 breast cancer cell lines and five lines from healthy tissue to EGF stimulation with or without perturbation with five kinase inhibitors. C_LIO_LISingle-cell signaling features and mechanistic signaling network models predicted drug sensitivity. C_LIO_LIMechanistic signaling network models deepen the understanding of drug resistance and sensitivity mechanisms. C_LIO_LIWe identify drug sensitivity-predictive genomic features via proxy signaling phenotypes. C_LI

systems biology↗

bioRxiv2019

The proteome of small urinary extracellular vesicles after kidney transplantation as an indicator of renal cellular biology and a source for markers predicting outcome

Kidney transplantation is the preferred renal replacement therapy available. Yet, the biological processes during and after kidney transplantation and how they translate into the overall functional graft outcome are insufficiently understood. Recent developments in the field of extracellular vesicle research allow the deeper exploitation of this non-invasive source. We separated small urinary extracellular vesicles (suEVs) throughout the course of living donor kidney transplantation. SuEVs were collected longitudinally from both the donor and the recipient in 22 living donor kidney transplantations. Unbiased proteomic analysis revealed specific temporal patterns of suEV proteins indicative of the cellular processes involved in the allografts response after transplantation with proteins playing a role in complement activation being among the most dynamically regulated components. Using a leave-one-out cross validation model, we identified potential prognostic markers of kidney function at 1 year after transplantation. One of the proteins identified - phosphoenol pyruvate carboxykinase (PCK2) - could be confirmed in an independent validation cohort of another 22 donor-recipient pairs using targeted mass spectrometry. This study sheds the light on early molecular processes during the course of kidney transplantation and shows the future potential of suEVs as a source of biomarkers in this setting. The data set is provided as a unique resource directly accessible through an online tool that allows dynamic interrogation of this first comprising suEV proteome atlas after kidney transplantation. One Sentence SummaryThis study represents the first atlas of the proteomic changes in small urinary extracellular vesicles throughout living donor kidney transplantation identifying PCK2 abundance as a biomarker for renal function 12 months after transplantation

molecular biology↗

bioRxiv2019

A natural variant of the sole pyruvate kinase of fission yeast lowers glycolytic flux triggering increased respiration and oxidative-stress resistance but decreased growth

Cells balance glycolysis with respiration to support their energetic and biosynthetic needs in different environmental or physiological contexts. With abundant glucose, many cells prefer to grow by aerobic glycolysis, or fermentation in yeast. Using 161 natural isolates of fission yeast, we investigated the genetic basis and phenotypic effects of the fermentation-respiration balance. The laboratory and a few other strains were more dependent on respiration. This trait was associated with a missense variant in a highly conserved region of Pyk1. Pyk1 is the single pyruvate kinase in fission yeast, while most organisms possess isoforms with different activity. This variant reduced Pyk1 activity and glycolytic flux. Replacing the low-activity pyk1 allele in the laboratory strain with the common high-activity allele was sufficient to increase fermentation and decrease respiration. This metabolic reprogramming triggered systems-level adaptations in the transcriptome and proteome, and in cellular phenotypes, including increased growth and chronological lifespan, but decreased resistance to oxidative stress. Thus, low Pyk1 activity provided no growth advantage but stress tolerance, despite increased respiration. The genetic tuning of glycolytic flux by a single-nucleotide change might reflect an adaptive trade-off in a species lacking pyruvate-kinase isoforms.

genetics↗

bioRxiv2019

A nutritional memory impairs survival, transcriptional and metabolic response to dietary restriction in old mice

Dietary restriction (DR) during adulthood can greatly extend lifespan and improve metabolic health in diverse species. However, whether DR in mammals is still effective when applied for the first time at old age remains elusive. Here, we conducted a late-life DR switch experiment employing 800 mice, by switching old animals from ad libitum (AL) to DR and vice versa. Strikingly, the switch from DR-to-AL acutely increased mortality, while the switch from AL-to-DR caused only a weak and gradual increase in survival, highlighting a memory of earlier nutrition. A significant association between fat preservation and survival response pointed to the white adipose tissue (WAT) as a potential memory source. Consistently, post-switch RNA-seq profiling in liver and WAT demonstrated that the transcriptional and metabolic program of chronic DR remained largely refractory to the AL-to-DR switch specifically in adipose tissue. Integration of lipidomics confirmed impaired membrane lipogenesis and limited mitochondrial copy number increase under late-life DR as functional consequences of this memory effect. Together, our results provide evidence for a nutritional memory as a limiting factor for DR-induced longevity and metabolic remodeling of WAT in mammals.

physiology↗

bioRxiv2019

Aging-associated changes in transcriptional elongation influence metazoan longevity.

Physiological homeostasis becomes compromised during aging, as a result of impairment of cellular processes, including transcription and RNA splicing. However, the molecular mechanisms leading to the loss of transcriptional fidelity are so far elusive, as are ways of preventing it. Here, we profiled and analyzed genome-wide, aging-related changes in transcriptional processes across different organisms: nematode worms, fruit flies, mice, rats and humans. The average transcriptional elongation speed (Pol-II speed) increased with age in all five species. Along with these changes in elongation speed we observed changes in splicing, including a reduction of unspliced transcripts and the formation of more circular RNAs. Two lifespan-extending interventions, dietary restriction and lowered insulin/Igf signaling, both reversed most of these aging-related changes. Genetic variants in Pol-II that reduced its speed in worms and flies increased their lifespan. Similarly, reducing Pol-II speed by overexpressing histone components, to counter age-associated changes in nucleosome positioning, also extended lifespan in flies and the division potential of human cells. Our findings uncover fundamental molecular mechanisms underlying animal aging and lifespan-extending interventions, and point to possible preventative measures.

genomics↗

bioRxiv2019

Integration of transcriptome, proteome and phosphoproteome data elucidates the genetic control of molecular networks

Genomic variation affects cellular networks by altering diverse molecular layers such as RNA levels, protein abundance, and post-translational protein modifications. However, it remains unclear how these different layers are affected by genetic polymorphisms and give rise to complex physiological phenotypes. To address these questions, we generated high-quality transcriptome, proteome, and phosphoproteome data for a panel of 112 genetically diverse yeast strains. While genetic effects on transcript abundances were generally transmitted to the protein level, we found a significant uncoupling of the transcript-protein relationship for certain protein classes, such as subunits of protein complexes. The additional phosphoproteomics data suggests that the same genetic locus often affects distinct sets of genes within each of these layers. In particular, QTLs tended to affect upstream regulatory proteins at the phosphorylation layer, whereas downstream pathway targets were typically affected at the transcript and protein abundance layers. Underscoring the importance of regulatory protein phosphorylation in linking genetic to phenotypic variation is the finding that the number of protein phosphosites associated with a given genetic locus was more predictive for its influence on cellular growth traits than the number of transcripts or proteins.\n\nThis study shows how multi-layered molecular networks mediate the effects of genomic variants to more complex physiological traits and highlights the important role of protein phosphorylation in mediating these effects.

systems biology↗

bioRxiv2019

Network-based imputation of dropouts in single-cell RNA sequencing data

Single-cell RNA sequencing (scRNA-seq) methods are typically unable to quantify the expression levels of all genes in a cell, creating a need for the computational prediction of missing values ( dropout imputation). Most existing dropout imputation methods are limited in the sense that they exclusively use the scRNA-seq dataset at hand and do not exploit external gene-gene relationship information. Further, it is unknown if all genes equally benefit from imputation or which imputation method works best for a given gene. Here, we show that a transcriptional regulatory network learned from external, independent gene expression data improves dropout imputation. Using a variety of human scRNA-seq datasets we demonstrate that our network-based approach outperforms published state-of-the-art methods. The network-based approach performs particularly well for lowly expressed genes, including cell-type-specific transcriptional regulators. Further, the cell-to-cell variation of 12.6% to 48.2% of the genes could not be adequately imputed by any of the methods that we tested. In those cases gene expression levels were best predicted by the mean expression across all cells, i.e. assuming no measurable expression variation between cells. These findings suggest that different imputation methods are optimal for different genes. We thus implemented an R-package called ADImpute (available via Bioconductor https://bioconductor.org/packages/release/bioc/html/ADImpute.html) that automatically determines the best imputation method for each gene in a dataset. Our work represents a paradigm shift by demonstrating that there is no single best imputation method. Instead, we propose that imputation should maximally exploit external information and be adapted to gene-specific features, such as expression level and expression variation across cells. Author summarySingle-cell RNA-sequencing (scRNA-seq) allows for gene expression to be quantified in individual cells and thus plays a critical role in revealing differences between cells within tissues and characterizing them in healthy and pathological conditions. Because scRNA-seq captures the RNA content of individual cells, lowly expressed genes, for which few RNA molecules are present in the cell, are easily missed. These events are called dropouts and considerably hinder analysis of the resulting data. In this work, we propose to make use of gene-gene relationships, learnt from external and more complete datasets, to estimate the true expression of genes that could not be quantified in a given cell. We show that this approach generally outperforms previously published methods, but also that different genes are better estimated with different methods. To allow the community to use our proposed method and combine it with existing ones, we created the R package ADImpute, available through Bioconductor.

bioinformatics↗

bioRxiv2019

The FLAME-accelerated Signalling Tool (FaST): A tool for facile parallelisation of flexible agent-based models of cell signalling

Agent-based modelling is particularly adept at modelling complex features of cell signalling pathways, where heterogeneity, stochastic and spatial effects are important, thus increasing our understanding of decision processes in biology in such scenarios. However, agent-based modelling often is computationally prohibitive to implement. Parallel computing, either on central processing units (CPUs) or graphical processing units (GPUs), can provide a means to improve computational feasibility of agent-based applications but generally requires specialist coding knowledge and extensive optimisation. In this paper, we address these challenges through the development and implementation of the FLAME-accelerated signalling tool (FaST), a software that permits easy creation and parallelisation of agent-based models of cell signalling, on CPUs or GPUs. FaST incorporates validated new agent-based methods, for accurate modelling of reaction kinetics and, as proof of concept, successfully converted an ordinary differential equation (ODE) model of apoptosis execution into an agent-based model. We finally parallelised this model through FaST on CPUs and GPUs resulting in an increase in performance of 5.8x (16 CPUs) and 53.9x respectively. The FaST takes advantage of the communicating X-machine approach used by FLAME and FLAME GPU to allow easy alteration or addition of functionality to parallel applications, but still includes inherent parallelisation optimisation. The FaST, therefore, represents a new and innovative tool to easily create and parallelise bespoke, robust, agent-based models of cell signalling.

systems biology↗

bioRxiv2019

A coevolved EDS1-SAG101-NRG1 module mediates cell death signaling by TIR-domain immune receptors

Plant intracellular nucleotide-binding/leucine-rich repeat (NLR) immune receptors are activated by pathogen effectors to trigger host defenses and cell death. Toll-Interleukin1-receptor (TIR)-domain NLRs (TNLs) converge on the Enhanced Disease Susceptibility1 (EDS1) family of lipase-like proteins for all resistance outputs. In Arabidopsis TNL immunity, AtEDS1 heterodimers with Phytoalexin Deficient4 (AtPAD4) transcriptionally boost basal defense pathways. AtEDS1 uses the same surface to interact with PAD4-related Senescence-Associated Gene101 (AtSAG101), but the role of AtEDS1-AtSAG101 heterodimers was unclear. We show that AtEDS1-AtSAG101 function together with AtNRG1 coiled-coil domain helper NLRs as a coevolved TNL cell death signaling module. AtEDS1-AtSAG101-AtNRG1 cell death activity is transferable to the solanaceous species, Nicotiana benthamiana, and cannot be substituted by AtEDS1-AtPAD4 with AtNRG1 or AtEDS1-AtSAG101 with endogenous NbNRG1. Analysis of EDS1-family evolutionary rate variation and heterodimer structure-guided phenotyping of AtEDS1 variants or AtPAD4-AtSAG101 chimeras identify closely aligned -helical coil surfaces in the AtEDS1-AtSAG101 partner C-terminal domains that are necessary for TNL cell death signaling. Our data suggest that TNL-triggered cell death and pathogen growth restriction are determined by distinctive features of EDS1-SAG101 and EDS1-PAD4 complexes and that these signaling machineries coevolved with further components within plant species or clades to regulate downstream pathways in TNL immunity.

plant biology↗

bioRxiv2019

HMGB1 as a rheostat of chromatin topology and RNA homeostasis on the path to senescence

Spatial organization and gene expression of mammalian chromosomes are maintained and regulated in conjunction with cell cycle progression. This is however disturbed once cells enter senescence and the highly abundant HMGB1 protein is depleted from senescent cell nuclei to act as an extracellular proinflammatory stimulus. Despite its physiological importance, we know little about the positioning of HMGB1 on chromatin or about its roles in the nucleus. To address this, we mapped HMGB1 binding genome-wide in different primary cells using a tailored protocol. We integrated ChIP-seq and Hi-C data with a graph theory approach to uncover HMGB1 demarcation of a subset of topologically-associating domains (TADs) that harbor genes required for paracrine senescence. Moreover, using sCLIP, knock-down and overexpression experiments, we now show that HMGB1 is a bona fide RNA-binding protein (RBP) bound to senescence-relevant mRNAs and affecting splicing. HMGB1 also has an interactome rich in RBPs, many of which are implicated in senescence regulation. The mRNAs of many of these RBPs are directly bound by HMGB1 and concertedly regulate the availability of SASP-relevant transcripts. Our findings highlight a broader than hitherto assumed role for HMGB1. It coordinates chromatin folding and RNA homeostasis as part of a feedforward loop controlling both cell-autonomous and paracrine senescence inside and outside of cells.

genomics↗

bioRxiv2018

Detection of epistatic interactions with Random Forest

In order to elucidate the influence of genetic factors on phenotype variation, non-additive genetic interactions (i.e., epistasis) have to be taken into account. However, there is a lack of methods that can reliably detect such interactions, especially for quantitative traits. Random Forest was previously recognized as a powerful tool to identify the genetic variants that regulate trait variation, mainly due to its ability to take epistasis into account. However, although it can account for interactions, it does not specifically detect them. Therefore, we propose three approaches that extract interactions from a Random Forest by testing for specific signatures that arise from interactions, which we termed paired selection frequency, split asymmetry, and selection asymmetry. Since they complement each other for different epistasis types, an ensemble method that combines the three approaches was also created. We evaluated our approaches on multiple simulated scenarios and two different real datasets from different Saccharomyces cerevisiae crosses. We compared them to the commonly used exhaustive pair-wise linear model approach, as well as several two-stage approaches, where loci are pre-selected prior to interaction testing. The Random Forest-based methods presented here generally outperformed the other methods at identifying meaningful genetic interactions both in simulated and real data. Further examination of the results for the simulated and real datasets established how interactions are extracted from the Random Forest, and explained the performance differences between the methods. Thus, the approaches presented here extend the applicability of Random Forest for the genetic mapping of biological traits.\n\nAuthor summaryThe genetic mechanisms underlying biological traits are often complex, involving the effects of multiple genetic variants. Interactions between these variants, also called epistasis, are also common. The machine learning algorithm Random Forest can be used to study genotype-phenotype relationships, by using genetic variants to predict the phenotype. One of Random Forests strengths is its ability to implicitly model interactions. However, Random Forest does not give any information about which predictors specifically interact, i.e. which variants are in epistasis.\n\nHere, we developed three approaches that identify interactions in a Random Forest. We demonstrated their ability to detect genetic interactions using simulations and real data from Saccharomyces cerevisiae. Our Random Forest-based methods generally outperformed several other commonly used approaches at detecting epistasis.\n\nThis study contributes to the long-standing problem of extracting information about the underlying model from a Random Forest. Since Random Forest has many applications outside of genetic association, this work represents a valuable contribution to not only genotype-phenotype mapping research, but also other scientific applications where interactions between predictors in a Random Forest might be of interest.

bioinformatics↗