Comparative analysis of four disease prediction models of Parkinson’s disease Nadella Kumudini 1 • Shaik Mohammad Naushad 2 • Balraj Alex Stanley 2 • Manoharan Niveditha 2 • Gunasekaran Sharmila 2 • Konda Kumaraswami 3 • Rupam Borghain 4 • Rukmini Mridula 4 • Vijay Kumar Kutala 3 Received: 8 May 2015 / Accepted: 26 September 2015 Ó Springer Science+Business Media New York 2015 Abstract Parkinson’s disease (PD) is a multi-factorial disorder with high-penetrant mutations accounting for small percentage of PD. Our previous studies demonstrated individual association of genetic variants in folate, xeno- biotic, and dopamine metabolic pathways with PD risk. The rational of the study was to develop a risk prediction model for PD using these genetic polymorphisms along with synuclein (SNCA) polymorphism. We have generated additive, multifactor dimensionality reduction (MDR), recursive partitioning (RP), and artificial neural network (ANN) models using 21 SNPs as inputs and disease out- come as output. The clinical utility of all these models was assessed by plotting receiver operating characteristics curves where in area under the curve (AUC) was used as an index of diagnostic utility of the model. The additive model was the simplest and exhibited an AUC of 0.72. The MDR model showed significant gene–gene interactions between SNCA, DRD4VNTR, and DRD2A polymorphisms. The RP model showed SHMT C1420T as important determi- nant of PD risk. This variant allele was found to be pro- tective and this protection was nullified by MTRR A66G. Inheritance of SHMT wild allele and SNCA intronic polymorphism was shown to increase the risk of PD. The ANN model showed higher diagnostic utility (AUC = 0.86) compared to all the models and was able to explain 56.6 % cases of sporadic PD. To conclude, the ANN model developed using SNPs in folate, xenobiotic, and dopamine pathways along with SNCA has higher clinical utility in predicting PD risk compared to other models. Keywords Parkinson’s disease Á Folate pathway Á Xenobiotic pathway Á Dopamine pathway Á Synuclein Á Artificial neural network Introduction Parkinson’s disease (PD) is characterized by the loss of dopaminergic neurons in the substantia nigra resulting in the disruption of putamen circuit thus manifesting in the form of tremors, rigidity, akinesia, and postural abnor- malities. Further, ‘‘Lewy bodies’’ are formed as a result of aggregation of a-synuclein-, synphylin-1-, and micro- tubule-associated proteins [1]. The familial PD accounts for mutations in a-synuclein, Parkin, and Ubiquitin carboxy terminal hydrogenase L1 (UGH-L1) [2]. The etiology of sporadic PD is complex involving several mechanisms such as oxidative stress [3], excitotoxicity [4], altered dopamine metabolism, and deficient detoxification of xenobiotic agents [5]. Earlier, we have reported gene–gene interactions between methionine synthase (MTRR) A66G and cytosolic serine hydrox- ymethyltransferase (cSHMT) C1420T in modulating PD risk [6]. MTRR A66G was associated with increased risk for PD, while cSHMT C1420T reduced the risk by mod- ulating homocysteine levels [6]. Further, we have also observed that cytochrome P450 (CYP)1A1 m1 and & Vijay Kumar Kutala vijaykutala@gmail.com 1 Department of Biotechnology, Jawaharlal Nehru Technological University-Hyderabad, Hyderabad, India 2 School of Chemical and Biotechnology, SASTRA University, Tirumalaisamudram, Thanjavur, India 3 Department of Clinical Pharmacology and Therapeutics, Nizam’s Institute of Medical Sciences, Panjagutta, Hyderabad, India 4 Department of Neurology, Nizam’s Institute of Medical Sciences, Panjagutta, Hyderabad, India 123 Mol Cell Biochem DOI 10.1007/s11010-015-2574-0