Biology subjects

Fernando, R. L.

Publications and source records attributed to Fernando, R. L..

4 recordsLinked to original sources

bioRxiv2022

Validation of the linear regression method to evaluate population accuracy and bias of predictions for non-linear models

BackgroundThe linear regression method (LR) was proposed to estimate population bias and accuracy of predictions, while addressing the limitations of commonly used cross-validation methods. The validity and behavior of the LR method have been provided and studied for linear model predictions but not for non-linear models. The objectives of this study were to 1) provide a mathematical proof for the validity of the LR method when predictions are based on conditional mean, 2) explore the behavior of the LR method in estimating bias and accuracy of predictions when the model fitted is different from the true model, and 3) provide guidelines on how to appropriately partition the data into training and validation such that the LR method can identify presence of bias and accuracy in predictions. ResultsWe present a mathematical proof for the validity of the LR method to estimate bias and accuracy of predictions based on the conditional mean, including for non-linear models. Using simulated data, we show that the LR method can accurately detect bias and estimate accuracy of predictions when an incorrect model is fitted when the data is partitioned such that the values of relevant predictor variables differ in the training and validation sets. But the LR method fails when the data are not partitioned in that manner. ConclusionsThe LR method was proven to be a valid method to evaluate the population bias and accuracy of predictions based on the conditional mean, regardless of whether it is a linear or non-linear function of the data. The ability of the LR method to detect bias and estimate accuracy of predictions when the model fitted is incorrect depends on how the data are partitioned. To appropriately test the predictive ability of a model using the LR method, the values of the relevant predictor variables need to be different between the training and validation sets.

genetics↗

bioRxiv2021

Interpretable Artificial Neural Networks incorporating Bayesian Alphabet Models for Genome-wide Prediction and Association Studies

In conventional linear models for whole-genome prediction and genome-wide association studies (GWAS), it is usually assumed that the relationship between genotypes and phenotypes is linear. Bayesian neural networks have been used to account for non-linearity such as complex genetic architectures. Here, we introduce a method named NN-Bayes, where "NN" stands for neural networks, and "Bayes" stands for Bayesian Alphabet models, including a collection of Bayesian regression models such as BayesA, BayesB, BayesC, Bayesian LASSO, and BayesR. NN-Bayes incorporates Bayesian Alphabet models into non-linear neural networks via hidden layers between SNPs and observed traits. Thus, NN-Bayes attempts to improve the performance of genome-wide prediction and GWAS by accommodating non-linear relationships between the hidden nodes and the observed trait, while maintaining genomic interpretability through the Bayesian regression models that connect the SNPs to the hidden nodes. For genomic interpretability, the posterior distribution of marker effects in NN-Bayes is inferred by Markov chain Monte Carlo (MCMC) approaches and used for inference of association through posterior inclusion probabilities (PIPs) and window posterior probability of association (WPPA). In simulation studies with dominance and epistatic effects, performance of NN-Bayes was significantly better than conventional linear models for both GWAS and whole-genome prediction, and the differences on prediction accuracy were substantial in magnitude. In real data analyses, for the soy dataset, NN-Bayes achieved significantly higher prediction accuracies than conventional linear models, and results from other four different species showed that NN-Bayes had similar prediction performance to linear models, which is potentially due to the small sample size. Our NN-Bayes is optimized for high-dimensional genomic data and implemented in an open-source package called "JWAS". NN-Bayes can lead to greater use of Bayesian neural networks to account for non-linear relationships due to its interpretability and computational performance.

genetics↗

bioRxiv2019

Exact Distribution of Linkage Disequilibrium in the Presence of Mutation, Selection or Minor Allele Frequency Filtering

Linkage disequilibrium (LD), often expressed in terms of the squared correlation (r2) between allelic values at two loci, is an important concept in many branches of genetics and genomics. Genetic drift and recombination have opposite effects on LD, and thus r2 will keep changing until the effects of these two forces are counterbalanced. Several approximations have been used to determine the expected value of r2 at equilibrium in the presence or absence of mutation. In this paper, we propose a probability-based approach to compute the exact distribution of allele frequencies at two loci in a finite population at any generation t conditional on the distribution at generation t - 1. As r2 is a function of this distribution of allele frequencies, this approach can be used to examine the distribution of r2 over generations as it approaches equilibrium. The exact distribution of LD from our method is used to describe, quantify and compare LD at different equilibria, including equilibrium in the absence or presence of mutation, selection, and filtering by minor allele frequency. We also propose a deterministic formula for expected LD in the presence of mutation at equilibrium based on the exact distribution of LD.

genetics↗

bioRxiv2016

The accuracy and bias of single-step genomic prediction for populationsunder selection

In single-step analyses, missing genotypes are explicitly or implicitly imputed, and this requires centering the observed genotypes, ideally using the mean of the unselected founders. If genotypes are only available on selected individuals, centering on the unselected founder mean is impossible. Here, computer simulation is used to study an alternative analysis that does not require centering genotypes but fits the mean {micro}g of unselected individuals as a fixed effect. To improve numerical properties of the analysis, centering the entire matrix of observed and imputed genotypes, using their sample means can be done in addition to fitting {micro}g. Starting with observed diplotypes from 721 cattle, a 5 generation population was simulated with sire selection to produce 40,000 individuals with phenotypes of which the 1,000 sires had genotypes. The next generation of 8,000 genotyped individuals was used for validation. Evaluations were undertaken: with (J) or without (N) {micro}g when marker covariates were not centered; and with (JC) or without (C) {micro}g when all marker covariates were centered. A pedigree based evaluation was less accurate than genomic analyses. Centering did not influence accuracy of genomic prediction, but fitting {micro}g did. Accuracies were improved when the panel comprised only QTL, models JC and J had accuracies of 99.2%; and models C and N had accuracies of 85.6%. When only markers were in the panel, the 4 models had accuracies of 63.9%. In panels that included causal variants, fitting {micro}g in the model improved accuracy, but had little impact when the panel contained only markers.

genomics↗