Introducing Unit Invariant Knee (UIK) As an Objective Choice for Elbow Point in Multivariate Data Analysis Techniques

Christopoulos, Demetris

doi:10.2139/ssrn.3043076

Cited by 27 publications

(24 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Model validation was performed by using the predictors and their importance values to predict the case/control phenotype in the test population. To determine which SNPs were important and worthy of further investigation, a scree plot was plotted and the second-order point of inflection was identified using the inflection R package (Christopoulos, 2016(Christopoulos, , 2017) (i.e., the "elbow method"). Predictors with importance values equal to or greater than the second-order point of inflection were defined as important SNPs and explored in downstream analyses if and only if the RF model was significantly more accurate at predicting phenotype in the test population than the non-information rate (i.e., the frequency of the more common phenotype).…”

Section: Random Forest Gwasmentioning

confidence: 99%

Genome-Wide Association Studies Reveal Susceptibility Loci for Noninfectious Claw Lesions in Holstein Dairy Cattle

et al. 2021

View full text Add to dashboard Cite

Sole ulcers (SUs) and white line disease (WLD) are two common noninfectious claw lesions (NICL) that arise due to a compromised horn production and are frequent causes of lameness in dairy cattle, imposing welfare and profitability concerns. Low to moderate heritability estimates of SU and WLD susceptibility indicate that genetic selection could reduce their prevalence. To identify the susceptibility loci for SU, WLD, SU and/or WLD, and any type of noninfectious claw lesion, genome-wide association studies (GWAS) were performed using generalized linear mixed model (GLMM) regression, chunk-based association testing (CBAT), and a random forest (RF) approach. Cows from five commercial dairies in California were classified as controls having no lameness records and ≥6 years old (n = 102) or cases having SU (n = 152), WLD (n = 117), SU and/or WLD (SU + WLD, n = 198), or any type of noninfectious claw lesion (n = 217). The top single nucleotide polymorphisms (SNPs) were defined as those passing the Bonferroni-corrected suggestive and significance thresholds in the GLMM analysis or those that a validated RF model considered important. Effects of the top SNPs were quantified using Bayesian estimation. Linkage disequilibrium (LD) blocks defined by the top SNPs were explored for candidate genes and previously identified, functionally relevant quantitative trait loci. The GLMM and CBAT approaches revealed the same regions of association on BTA8 for SU and BTA13 common to WLD, SU + WLD, and NICL. These SNPs had effects significantly different from zero, and the LD blocks they defined explained a significant amount of phenotypic variance for each dataset (6.1–8.1%, p < 0.05), indicating the small but notable contribution of these regions to susceptibility. These regions contained candidate genes involved in wound healing, skin lesions, bone growth and mineralization, adipose tissue, and keratinization. The LD block defined by the most significant SNP on BTA8 for SU included a SNP previously associated with SU. The RF models were overfitted, indicating that the SNP effects were very small, thereby preventing meaningful interpretation of SNPs and any downstream analyses. These findings suggested that variants associated with various physiological systems may contribute to susceptibility for NICL, demonstrating the complexity of genetic predisposition.

show abstract

Section: Random Forest Gwasmentioning

confidence: 99%

Genome-Wide Association Studies Reveal Susceptibility Loci for Noninfectious Claw Lesions in Holstein Dairy Cattle

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The scree plot for a SBM graph generated using the parameters of our surrogate model ( 1), ( 2) is shown in Figure 1. Estimating the elbow point using the unit-invariant knee method (Christopoulos, 2016) yields an optimum value of d = 4. This choice of d − 4 is also consistent if we instead use an alternative method (Satopaa, Albrecht, Irwin, & Raghavan, 2011) of estimating the distance from each point in the scree plot to a line joining the first and last points of the plot, and then selecting the elbow point where this distance is the largest.…”

Section: Embedding In a Lower Dimensionmentioning

confidence: 99%

Neuronal classification from network connectivity via adjacency spectral embedding

Mehta

Goldin

Marchette

et al. 2021

Network Neuroscience

View full text Add to dashboard Cite

This work presents a novel strategy for classifying neurons, represented by nodes of a directed graph, based on their circuitry (edge connectivity). We assume a stochastic block model (SBM) in which neurons belong together if they connect to neurons of other groups according to the same probability distributions. Following adjacency spectral embedding of the SBM graph, we derive the number of classes and assign each neuron to a class with a Gaussian mixture model-based expectation-maximization (EM) clustering algorithm. To improve accuracy, we introduce a simple variation using random hierarchical agglomerative clustering to initialize the EM algorithm and picking the best solution over multiple EM restarts. We test this procedure on a large (≈212–215 neurons), sparse, biologically inspired connectome with eight neuron classes. The simulation results demonstrate that the proposed approach is broadly stable to the choice of embedding dimension, and scales extremely well as the number of neurons in the network increases. Clustering accuracy is robust to variations in model parameters and highly tolerant to simulated experimental noise, achieving perfect classifications with up to 40% of swapped edges. Thus, this approach may be useful to analyze and interpret large-scale brain connectomics data in terms of underlying cellular components.

show abstract

“…To determine the inflection point marking the transition between the first and the second cluster of pest species, we fitted a GAM to the ordered turning-points for each cut-off threshold (0.4-0.6) and identified the inflection point temperature with the function ese in the 'inflection' R-package (version 1.3.5; Christopoulos, 2016).…”

Section: Analysis Of Inflection Points Under Climate Changementioning

confidence: 99%

Inflection point in climatic suitability of insect pest species in Europe suggests non‐linear responses to climate change

Grünig

Calanca

Mazzi

et al. 2020

Global Change Biology

View full text Add to dashboard Cite

Climate change and globalization affect the suitable conditions for agricultural crops and insect pests, threatening future food security. It remains unknown whether shifts in species' climatic suitability will be linear or rather non-linear, with crop exposure to pests suddenly increasing when a critical temperature threshold is crossed. Moreover, uncertainty of forecasts can arise because of the modelling approach based either on species distribution data or on physiological measurements. Here, we compared the predictions of two modelling approaches (physiological models and species distribution models) for forecasting the potential distribution of agricultural insect pests in Europe. Despite conceptual differences, we found good agreement overall between the two approaches. We further identified a potential regime change in pest pressure along a temperature gradient. With both modelling approaches, we found an inflection point in the number of pest species with suitable climatic conditions around a minimum temperature of the coldest month of −3°C. Our results could help decision-makers anticipate the onset of rising pest pressure and provide support for intensifying surveillance measures, particularly in regions where temperatures are already beyond the inflection point. K E Y W O R D S agricultural crop, climate change, insect pest, physiological model, species distribution model, temperature threshold | 6339 GRÜNIG et al.

show abstract

Introducing Unit Invariant Knee (UIK) As an Objective Choice for Elbow Point in Multivariate Data Analysis Techniques

Cited by 27 publications

References 8 publications

Genome-Wide Association Studies Reveal Susceptibility Loci for Noninfectious Claw Lesions in Holstein Dairy Cattle

Genome-Wide Association Studies Reveal Susceptibility Loci for Noninfectious Claw Lesions in Holstein Dairy Cattle

Neuronal classification from network connectivity via adjacency spectral embedding

Inflection point in climatic suitability of insect pest species in Europe suggests non‐linear responses to climate change

Contact Info

Product

Resources

About