ABSTRACT
Multifactorial disorders are diseases whose underlying causes may be due to a combination of factors, including genetics, environmental influences, and lifestyle. Deconvoluting the factors responsible for such issues can be tremendously challenging. However, evolutionary algorithms (EAs) are particularly well suited to address this class of problem. EAs may be used to generate populations of solutions to carry out ensemble-based analyses of mathematical models that can determine which factors may have the largest impact on a particular disease state. In this work, a commonly utilized mathematical model for bone physiology was used to explore the underlying mechanism responsible for secondary hyperparathyroidism due to renal failure. Using a genetic algorithm (GA), a population of parameter sets for the bone model were generated based on a clinical data set. The average result of the best five GA-based models was compared to the standard bone model prediction as reported in the literature. The GA-based models manipulated several parameters simultaneously while the standard literature model manipulated only the glomerular filtration rate (GFR) parameter. The GA model resulted in a superior fit of the clinical data. It further indicated that six of the 108 parameters were significantly changed from what would be expected of a healthy individual. The GFR parameter was among the six significantly changed parameters. The remaining significantly changed parameters involved phosphate imbalances. Interestingly, hyperphosphatemia has been recognized as one of the primary causes of secondary hyperparathyroidism. These results would appear to corroborate the effectiveness of the GA in better understanding multifactorial disorders.
- Eddy J Bautista and Ranjan Srivastava. 2013. Leveraging ensemble information of evolving populations in genetic algorithms to identify incomplete metabolic pathways. In Proceedings of the 15th annual conference companion on Genetic and evolutionary computation. ACM, 39--40. Google ScholarDigital Library
- Eddy J Bautista, Joseph Zinski, Steven M Szczepanek, Erik L Johnson, Edan R Tulman, Wei-Mei Ching, Steven J Geary, and Ranjan Srivastava. 2013. Semiautomated curation of metabolic models via flux balance analysis: a case study with Mycoplasma gallisepticum. PLoS Computational Biology 9, 9 (2013), e1003208.Google ScholarCross Ref
- I Lobo. 2008. Multifactorial inheritance and genetic disease. Nat Educ 1, 1 (2008).Google Scholar
- Mark C Peterson and Matthew M Riggs. 2010. A physiologically based mathematical model of integrated calcium homeostasis and bone remodeling. Bone 46, 1 (2010), 49--63.Google ScholarCross Ref
- AA Portale and RC Morris Jr. 1991. Pathogenesis of secondary hyperparathyroidism in chronic renal insufficiency. Mineral and electrolyte metabolism 17, 4 (1991), 211--220.Google Scholar
- Wissam Saliba and Boutros El-Haddad. 2009. Secondary hyperparathyroidism: pathophysiology and treatment. The Journal of the American Board of Family Medicine 22, 5 (2009), 574--581.Google ScholarCross Ref
Index Terms
- Using ensemble modeling to determine causes of multifactorial disorders
Recommendations
Ensemble of niching algorithms
Although niching algorithms have been investigated for almost four decades as effective procedures to obtain several good and diverse solutions of an optimization problem, no effort has been reported on combining different niching algorithms to form an ...
Differential evolution algorithm with ensemble of parameters and mutation strategies
Differential evolution (DE) has attracted much attention recently as an effective approach for solving numerical optimization problems. However, the performance of DE is sensitive to the choice of the mutation strategy and associated control parameters. ...
Using holey fitness landscapes to counteract premature convergence in evolutionary algorithms
GECCO '08: Proceedings of the 10th annual conference companion on Genetic and evolutionary computationPremature convergence is a persisting problem in evolutionary optimisation, in particular - genetic algorithms. While a number of methods exist to approach this issue, they usually require problem specific calibration or only partially resolve the issue,...
Comments