Supervised machine learning techniques for the classification of metabolic disorders in newborns

Abstract

Six machine learning techniques have been investigated for their classification accuracy focusing on two metabolic disorders, phenylketo nuria (PKU) and medium-chain acyl-CoA dehydrogenase deficiency (MCADD). Logistic regression analysis led to superior classification rules (sensitivity >96.8%, specificity >99.98%) compared to all investigated algorithms. Including novel constellations of metabolites into the models, the positive predictive value could be strongly increased (PKU 71.9% versus 16.2%, MCADD 88.4% versus 54.6% compared to the established diagnostic markers). Our results clearly prove that the mined data confirm the known and indicate some novel metabolic patterns which may contribute to a better understanding of newborn metabolism.

Targeted Metabolomics for Clinical Biomarker Discovery in Multifactorial Diseases

“…It also supports the design of tailored kinetic models of human-specific metabolic pathways including detailed knowledge about all metabolic reactions concerned. Besides statistical model building and data mining-based approaches (Baumgartner et al, 2004;Baumgartner et al, 2005;Baumgartner & Graber, 2008), computational systems biology is essential to combine knowledge of human physiology and pathology starting from genomics, molecular biology and the environment through the levels of cells, tissues, and organs all the way up to integrated systems behaviour. Applying systems biology approaches within the context of human health and disease will definitely gain new insights.…”

Biomarker Discovery, Disease Classification, and Similarity Query Processing on High-Throughput MS/MS Data of Inborn Errors of Metabolism

“…This unbalance of class size is necessary to avoid an overestimation of TP rates in the control class. 21 Because TPc 2 rates (specificities) ≥ 99.6% were computed in all study experiments, the performance measure TP* is predominantly determined by the value of the TPc 1 rate (sensitivity) describing the fraction of correctly classified diseased subjects. TP* is thus more sensitive than overall classification accuracy (= correctly classified subjects in both classes/all subjects), which does not reflect the unbalancedness of classes.…”

“…In this study, we compared several popular methods-that is, LRA, k-NN, naive Bayes, support vector machines (SVM), and artificial neural networks (ANN), which are used for classifying metabolomic/proteomic data. 21,[23][24][25][26] For general information on classification algorithms, see, for example, Mitchell, 27 Cristianini and Shawe-Taylor, 28 Shawe-Taylor and Cristianini, 29 Gelman et al, 30 and Raudys. 31 We examined classification accuracy of classifiers with respect to (w.r.t.)…”

Subspace Selection for Clustering High-Dimensional Data

“…Phe-Glu-ArgSuc). Alterations of our best ranked single metabolites correspond well to the abnormal metabolism of PKU [5]. We compared SURF-ING findings with results using PCA.…”