Received on October 17, 2006
Accepted on April 30, 2007

Informatics and Statistics

Feature Construction Can Improve Diagnostic Criteria for High-Dimensional Metabolic Data in Newborn Screening for Medium-Chain Acyl-CoA Dehydrogenase Deficiency

¹ Division of Metabolic Diseases, Department of General Pediatrics, University Children's Hospital, Heidelberg, Germany.
² Department of Pediatrics and Institute of Clinical Chemistry, Hamburg University Medical Center, Hamburg, Germany.
³ Department of Medical Informatics, University of Heidelberg Medical Center, Heidelberg, Germany

^* To whom correspondence should be addressed. E-mail: Sirikit.Ho{at}med.uni-heidelberg.de.

Background: In newborn screening with tandem mass spectrometry, multiple intermediary metabolites are quantified in a single analytical run for the diagnosis of fatty-acid oxidation disorders, organic acidurias, and aminoacidurias. Published diagnostic criteria for these disorders normally incorporate a primary metabolic marker combined with secondary markers, often analyte ratios, for which the markers have been chosen to reflect metabolic pathway deviations.

Methods: We applied a procedure to extract new markers and diagnostic criteria for newborn screening to the data of newborns with confirmed medium-chain acyl-CoA dehydrogenase deficiency (MCADD) and a control group from the newborn screening program, Heidelberg, Germany. We validated the results with external data of the screening center in Hamburg, Germany. We extracted new markers by performing a systematic search for analyte combinations (features) with high discriminatory performance for MCADD. To select feature thresholds, we applied automated procedures to separate controls and cases on the basis of the feature values. Finally, we built classifiers from these new markers to serve as diagnostic criteria in screening for MCADD.

Results: On the basis of ² scores, we identified 700 of >628 000 new analyte combinations with superior discriminatory performance compared with the best published combinations. Classifiers built with the new features achieved diagnostic sensitivities and specificities approaching 100%.

Conclusion: Feature construction methods provide ways to disclose information hidden in the set of measured analytes. Other diagnostic tasks based on high-dimensional metabolic data might also profit from this approach.