Using network analysis to study behavioural phenotypes: an example using domestic dogs

Goold, Conor; Vas, Judit; Olsen, Christine; Newberry, Ruth C.

doi:10.1098/rsos.160268

Cited by 6 publications

(10 citation statements)

References 85 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The null hypothesis model (M1) was strongly rejected (ΔẼBIC = 56.08), as was our initial two factor hypothesis (M2; ΔẼBIC = 42.97). Both the oblique model (M3; ΔẼBIC = 41.84) and the two-factor solution with a residual zero-order correlation (M4; ΔẼBIC = 31.13) received little to no support compared to the EGA+GNM solutions, supporting the claim that partial correlation networks are generally more informative than zero-order correlations (Costantini et al, 2015;Epskamp & Fried, 2016;Goold et al, 2016). The pure FA models were also highly unstable compared to the GNM (M2-M7 admissible sample range:…”

Section: Resultsmentioning

confidence: 96%

“…Generate and test pairwise partial correlation network models Costantini et al (2015), Goold et al (2016) and Epskamp and Fried (2016) Exploratory graph analysis (EGA)…”

Section: Factor Analysismentioning

confidence: 99%

“…A GGM is therefore a network of partial correlations, where each correlation controls for the effect of all other correlations in the network. In Figure 1b, In contrast to the causal effects indicated by directed edges in FA and SEM, the undirected edges in a GGM can be conceptualized as pairwise interactions between observed traits, which may represent both directional and reciprocal causal processes (Costantini et al, 2015;Epskamp & Fried, 2016;Goold et al, 2016). Importantly, the undirected edges encoded by a GGM completely specify a unique set of conditional independence relationships, so that there are no equivalent GGMs possible given a particular dataset and estimation procedure.…”

Section: Generalized Network Modellingmentioning

confidence: 99%

“…For example, human intelligence is often explained with a single latent causal factor 'g', but alternative models positing developmental feedback between distinct cognitive processes can predict statistically equivalent performance outcomes (Van der Maas et al, 2006). An overemphasis upon unobserved causes may therefore obscure direct causal interactions between behavioural traits (Cramer et al, 2012;Goold et al, 2016;Schmittmann et al, 2013). This general problem of model equivalence also underscores the importance of theoretically informed model testing (Keith, Caemmerer, & Reynolds, 2016;Skrondal & Rabe-Hesketh, 2004) and subsequent empirical investigation to uncover the potential causal mechanisms represented by latent factors (Shipley, 2016).…”

Section: Exploratory Graph Analysismentioning

confidence: 99%

“…Although useful for generating models of unobserved common causes for multiple traits, FA is limited in its capacity to capture processes of pairwise interaction between behaviours (Cramer et al, 2012;Goold, Vas, Olsen, & Newberry, 2016;Schmittmann et al, 2013). Pairwise interactions here refer to direct associations between behaviours or states closely proxied by particular behavioural measures (e.g., social dominance, Favati, Leimar, Radesäter, & Løvlie, 2014) that reflect directional or reciprocal causes (e.g., energetic trade-offs or positive feedback processes).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

TheEGA+GNMframework: An integrative approach to modelling behavioural syndromes

et al. 2018

View full text Add to dashboard Cite

Behavioural syndromes refer to correlated suites of behavioural traits exhibiting consistent among‐individual variation, i.e. personality. Factor analysis (FA) is currently the dominant method for modelling behavioural syndromes in humans and animals. Although FA is useful for inferring the latent causes underlying trait correlations, it does not account for the pairwise behavioural interactions that also contribute to syndrome structure. Given that latent factors and pairwise interactions are likely ubiquitous causes of trait covariation, both should be modelled simultaneously. Currently, however, behavioural ecologists lack an integrative framework for describing and inferring such behavioural syndromes. Generalized network modelling (GNM), representing an integration of FA and Gaussian graphical modelling (GGM), meets this challenge. We provide a theoretical introduction to GNM as well as a method for detecting latent factors in GGMs called exploratory graph analysis (EGA). We then propose the novel EGA+GNM framework for modelling multiple sources of trait correlations and ensuring more robust causal inferences. To empirically demonstrate the utility of this framework, we compare models derived from EGA+GNM and FA using observational measures of social and arousal behaviour in common marmosets Callithrix jacchus. Using information‐theoretic model comparison, we find support for EGA+GNM models compared to models generated by FA. Two EGA+GNM models suggest that while latent factors contribute to the emergence of clustered sociability and arousal behaviours, correlations among these traits may also be partially explained by pairwise interactions. Additionally, these behavioural clusters are hypothesized to be causally linked by a positive pairwise interaction between allogrooming and activity level. These results support our claim that EGA+GNM provides a superior and integrative framework for describing behavioural syndromes. Consequently, by simultaneously modelling both latent factors and pairwise interactions, behavioural ecologists can better understand the evolutionary causes and consequences of animal personality. A formal overview of the EGA+GNM framework and a R tutorial demonstrating its application are provided in the electronic .

show abstract

Section: Resultsmentioning

confidence: 96%

“…Generate and test pairwise partial correlation network models Costantini et al (2015), Goold et al (2016) and Epskamp and Fried (2016) Exploratory graph analysis (EGA)…”

Section: Factor Analysismentioning

confidence: 99%

Section: Generalized Network Modellingmentioning

confidence: 99%

Section: Exploratory Graph Analysismentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

TheEGA+GNMframework: An integrative approach to modelling behavioural syndromes

et al. 2018

View full text Add to dashboard Cite

show abstract

Network analysis of canine brain morphometry links tumour risk to oestrogen deficiency and accelerated brain ageing

Rzechorzek

Saunders

Hiscox

et al. 2019

Sci Rep

View full text Add to dashboard Cite

Structural ‘brain age’ is a valuable but complex biomarker for several brain disorders. The dog is an unrivalled comparator for neurological disease modeling, however canine brain morphometric diversity creates computational and statistical challenges. Using a data-driven approach, we explored complex interactions between patient metadata, brain morphometry, and neurological disease. Twenty-four morphometric parameters measured from 286 canine brain magnetic resonance imaging scans were combined with clinical parameters to generate 9,438 data points. Network analysis was used to cluster patients according to their brain morphometry profiles. An ‘aged-brain’ profile, defined by a small brain width and volume combined with ventriculomegaly, was revealed in the Boxer breed. Key features of this profile were paralleled in neutered female dogs which, relative to un-neutered females, had an 11-fold greater risk of developing brain tumours. Boxer dog and geriatric dog groups were both enriched for brain tumour diagnoses, despite a lack of geriatric Boxers within the cohort. Our findings suggest that advanced brain ageing enhances brain tumour risk in dogs and may be influenced by oestrogen deficiency—a risk factor for dementia and brain tumours in humans. Morphometric features of brain ageing in dogs, like humans, might better predict neurological disease risk than patient chronological age.

show abstract

Network analysis of canine brain morphometry links tumour risk to oestrogen deficiency and accelerated brain ageing

Rzechorzek

Saunders

Hiscox

et al. 2018

Preprint

View full text Add to dashboard Cite

Background: Structural 'brain age' is a valuable but complex biomarker for several brain disorders. The dog is an unrivalled comparator for neurological disease modeling, however brain phenotypic diversity among pedigrees creates computational and statistical challenges.Methods: We applied unbiased network correlation analysis in dogs to explore complex interactions between brain morphometrics, patient metadata, and neurological disease.Twenty-four parameters measured from each of 286 brain magnetic resonance imaging scans 2 generated 9,438 data points that were used to cluster canine patients according to their brain morphometry profiles. The network was then explored for statistically significant enrichments within breed, sex, age, and diagnostic categories.Findings: Morphometric comparisons revealed an advanced 'aged-brain' profile in the Boxer breed, consisting of a small brain length, width, and volume, combined with ventriculomegaly. Key features of this profile were paralleled in neutered female dogs which, relative to un-neutered females, had an 11-fold greater risk of developing primary brain tumours. Enrichment analysis confirmed that Boxers and geriatric individuals were enriched for brain tumour diagnoses, despite a lack of geriatric Boxers within the cohort.Interpretation: These findings suggest that accelerated brain ageing might contribute to tumour risk in Boxers and may be influenced by oestrogen deficiency -a risk factor for dementia and brain tumours in humans. We propose that morphometric features of brain ageing in dogs, like humans, might better predict neurological disease risk than a patient's chronological age. Funding: Wellcome Trust Integrated Training Fellowship for Veterinarians (096409/Z/11/Z to N.M.R) and an MSD Animal Health Connect Bursary (to O.M.S.). Research in context Evidence before this studyOur ageing human population has increased the prevalence of chronic brain disorders, placing an unsustainable burden on healthcare systems worldwide. Contributing to this problem is our limited understanding of how chronic brain disorders begin, how they might Added value of this studyWith an unconventional data-driven approach, we have identified an advanced 'brain age' in neutered female dogs and the Boxer breed that is associated with brain tumour risk. To our knowledge, this is the first study to link brain tumour risk to both accelerated brain ageing and oestrogen loss. Moreover, we have demonstrated that neutering status affects brain shape in dogs. This study outlines new collaborative opportunities -through comparative biology -to understand the influence of chromosomal and hormonal sex on brain structure, brain ageing, and brain tumour development. Our data can also be incorporated into larger canine databases to help extract non-invasive image-based biomarkers for neurological disease.Furthermore, our unique network analysis approach to handle the combined complexity within clinical and brain imaging datasets is immediately relevant to human patient studies. Implications of all...

show abstract

Using network analysis to study behavioural phenotypes: an example using domestic dogs

Cited by 6 publications

References 85 publications

TheEGA+GNMframework: An integrative approach to modelling behavioural syndromes

TheEGA+GNMframework: An integrative approach to modelling behavioural syndromes

Network analysis of canine brain morphometry links tumour risk to oestrogen deficiency and accelerated brain ageing

Network analysis of canine brain morphometry links tumour risk to oestrogen deficiency and accelerated brain ageing

Contact Info

Product

Resources

About