Abstract
Studies have been performed to describe the significance of genetic polymorphisms in complications associated with disturbed perinatal adaptation. Due to the large number of interacting factors, the results of classic statistical methods are often inconsistent. The random forest technique (RFT) is a robust nonparametric statistical approach that overcomes this problem through the calculation of the importance of each factor. We used RFT to reanalyze the importance of 24 genetic polymorphisms in the classification of preterm infants (birth weight, 680–1460 g, n = 100) to affected and unaffected groups according to the presence of acute perinatal complications. The accuracy of classification was between 0.5 and 0.8 for each complication when only birth data were considered. However, when genetic polymorphisms with the highest importance scores (ISs) were included in the analysis, the accuracy of classification according overall morbidity, necrotizing enterocolitis (NEC), acute renal failure (ARF), infant respiratory distress syndrome (IRDS), cardiac failure (CF), and patent ductus arteriosus (PDA) improved from 0.69, 0.60, 0.70, 0.72, 0.68, and 0.57 to 0.77, 0.70, 0.76, 0.77, 0.76, and 0.64, respectively. Our findings suggest that genetic polymorphisms identified by RFT as predictors may improve the risk assessment of preterm infants. RFT is a suitable tool to develop risk factor patterns in this population.
Similar content being viewed by others
Log in or create a free account to read this content
Gain free access to this article, as well as selected content from this journal and more on nature.com
or
Abbreviations
- ARF:
-
Acute renal failure;
- CF:
-
cardiac failure;
- IRDS:
-
infant respiratory distress syndrome
- IS:
-
importance score;
- IVH:
-
intraventricular bleeding;
- NEC:
-
necrotizing enterocolitis;
- PDA:
-
patent ductus arteriosus;
- RFT:
-
random forest technique
References
Treszl A, Tulassay T, Vásárhelyi B 2006 Genetic basis for necrotizing enterocolitis—risk factors and their relations to genetic polymorphisms. Front Biosci 11: 570–580
Parton LA, Strassberg SS, Qian D, Galvin-Parton PA, Cristea IA 2006 The genetic basis for bronchopulmonary dysplasia. Front Biosci 11: 1854–1860
Vásárhelyi B, Tóth-Heyn P, Treszl A, Tulassay T 2005 Genetic polymorphisms and risk for acute renal failure in preterm neonates. Pediatr Nephrol 20: 132–135
Baier RJ, Loggins J, Yanamandra K 2006 IL-10, IL-6 and CD14 polymorphisms and sepsis outcome in ventilated very low birth weight infants. BMC Med 4: 10–
Treszl A, Szabó M, Dunai G, Nobilis A, Kocsis I, Machay T, Tulassay T, Vásárhelyi B 2003 Angiotensin II type 1 receptor A1166C polymorphism and prophylactic indomethacin treatment induced ductus arteriosus closure in very low birth weight neonates. Pediatr Res 54: 753–755
Baier RJ 2006 Genetics of perinatal brain injury in the preterm infant. Front Biosci 11: 1371–1387
Breiman L 2001 Random forests. Mach Learn 45: 5–32
Bureau A, Dupuis J, Falls K, Lunetta KL, Hayward B, Keith TP, Van Eerdewegh P 2005 Identifying SNPs predictive of phenotype using random forests. Genet Epidemiol 28: 171–182
Breiman L, Cutler A 2004 Random forests. Available at: http://www.stat.berkeley.edu/~breiman/RandomForests/cc_software.htm (accessed June 20, 2007)
R Development Core Team 2006 R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available at: http://www.R-project.org (accessed June 20, 2007)
Hothorn T, Hornik K, Zeileis A 2006 Party: a laboratory for recursive part(y)itioning (R package version 0.9-0) 2006 Available at: http://CRAN.R-project.org/ (accessed June 20, 2007)
Nobilis A, Szabó M, Kocsis I, Sulyok E, Tulassay T, Vásárhelyi B 2006 Angiotensin-converting enzyme DD genotype is preventive against circulatory failure in very low-birth-weight neonates. Acta Paediatr 95: 747–750
Treszl A, Kocsis I, Szathmari M, Schuler A, Heninger E, Tulassay T, Vásárhelyi B 2003 Genetic variants of TNF-α, IL-1beta, IL-4 receptor α-chain, IL-6 and IL-10 genes are not risk factors for sepsis in low-birth-weight infants. Biol Neonate 83: 241–245
Ahrens P, Kattner E, Kohler B, Hartel C, Seidenberg J, Segerer H, Moller J, Gopel W, Genetic Factors in Neonatology Study Group 2004 Mutations of genes involved in the innate immune system as predictors of sepsis in very low birth weight infants. Pediatr Res 55: 652–656
Harding D, Dhamrait S, Marlow N, Whitelaw A, Gupta S, Humphries S, Montgomery H 2003 Angiotensin-converting enzyme DD genotype is associated with worse perinatal cardiorespiratory adaptation in preterm infants. J Pediatr 143: 746–749
Bokodi G, Derzbach L, Vásárhelyi B 2006 Re: Deletion allele of angiotensin-converting enzyme. J Pediatr 149: 579–580
Kazzi SN, Quasney MW 2005 Deletion allele of angiotensin-converting enzyme is associated with increased risk and severity of bronchopulmonary dysplasia. J Pediatr 147: 818–822
Adcock K, Hedberg C, Loggins J, Kruger TE, Baier RJ 2003 The TNF-α −308, MCP-1-2518 and TGF-beta1 +915 polymorphisms are not associated with the development of chronic lung disease in very low birth weight infants. Genes Immun 4: 420–426
Bokodi G, Treszl A, Derzbach L, Balogh Á, Vásárhelyi B 2005 The association of the carrier state of the tumor necrosis factor-α (TNFα) (-308)A allele with the duration of oxygen supplementation in preterm neonates. Eur Cytokine Netw 16: 78–80
Diaz-Uriarte R, Alvarez de Andres S 2006 Gene selection and classification of microarray data using random forest. BMC Bioinform 7: 3
Heidema AG, Boer JM, Nagelkerke N, Mariman EC, van der A DL, Feskens EJ 2006 The challenge for genetic epidemiologists: how to analyze large numbers of SNPs in relation to complex diseases. BMC Genet 7: 23
Acknowledgements
The data analyzed were partly derived from the published work of former Ph.D. students in the authors' laboratory. The authors thank Ilona Banyasz, Geza Bokodi, Laszlo Derzbach, Erika Heninger, Andrea Fekete, Istvan Kocsis, Andras Nobilis, and Adam Vannay for their contribution.
Author information
Authors and Affiliations
Corresponding author
Additional information
This study was supported by grants OTKA T046086, GVOP 3.1.1-2004/300 and Philips Hungary Ltd. A.T and B.V. are recipients of a Bolyai Fellowship.
Rights and permissions
About this article
Cite this article
Treszl, A., Kaposi, A., Hajdú, J. et al. The Extent to Which Genotype Information May Add to the Prediction of Disturbed Perinatal Adaptation: None, Minor, or Major?. Pediatr Res 62, 610–614 (2007). https://doi.org/10.1203/PDR.0b013e318155a0e1
Received:
Accepted:
Issue date:
DOI: https://doi.org/10.1203/PDR.0b013e318155a0e1
