Application of machine learning tools and integrated OMICS for screening and diagnosis of inborn errors of metabolism.

INTRODUCTION: Tandem mass spectrometry (TMS) has emerged an important screening tool for various metabolic disorders in newborns. However, there is inherent risk of false positive outcomes. Objective To establish analyte-specific cutoffs in TMS by integrating metabolomics and genomics data to avoid false positivity and false negativity and improve its clinical utility. METHODS: TMS was performed on 572 healthy and 3000 referred newborns. Urine organic acid analysis identified 23 types of inborn errors in 99 referred newborns. Whole exome sequencing was performed in 30 positive cases. The impact of physiological changes such as age, gender, and birthweight on various analytes was explored in healthy newborns. Machine learning tools were used to integrate demographic data with metabolomics and genomics data to establish disease-specific cut-offs; identify primary and secondary markers; build classification and regression trees (CART) for better differential diagnosis; for pathway modeling. RESULTS: This integration helped in differentiating B12 deficiency from methylmalonic acidemia (MMA) and propionic acidemia (Phi coefficient=0.93); differentiating transient tyrosinemia from tyrosinemia type 1 (Phi coefficient=1.00); getting clues about the possible molecular defect in MMA to initiate appropriate intervention (Phi coefficient=1.00); to link pathogenicity scores with metabolomics profile in tyrosinemia (r2=0.92). CART model helped in establishing differential diagnosis of urea cycle disorders (Phi coefficient=1.00). CONCLUSION: Calibrated cut-offs of different analytes in TMS and machine learning-based establishment of disease-specific thresholds of these markers through integrated OMICS have helped in improved differential diagnosis with significant reduction of the false positivity and false negativity rates.

Application of machine learning algorithms for the differential diagnosis of peroxisomal disorders.

We have established diagnostic thresholds of very long-chain fatty acids (VLCFA) for the differential diagnosis of peroxisomal disorders using the machine learning tools. The plasma samples of 131 controls and 90 cases were tested for VLCFA using gas chromatography-mass spectrometry following stable isotope dilution. These data were used to construct association rules and for recursive partitioning. The C26/22 in healthy controls ranged between 0.008 and 0.01. The C26 levels between 1.61 and 3.34 µmol/l and C26/C22 between 0.05 and 0.10 are diagnostic of X-linked adrenoleukodystrophy (X-ALD). Very high levels of C26 (>3.34 µmol/l) and C26/C22 ratio (>0.10) are diagnostic of Zellweger syndrome (ZS). Significant elevation of phytanic acid was observed in Refsum (t = 6.14, P < 0.0001) and Rhizomelic chondrodysplasia punctata (RCDP) (t = 16.72, P < 0.0001). The C26/C22 ratio is slightly elevated in RCDP (t = 2.58, P = 0.01) while no such elevation was observed in Refsum disease (t = 0.86, P = 0.39). The developed algorithm exhibited greater clinical utility (AUC: 0.99-1.00) in differentiating X-ALD, ZS and healthy controls. The algorithm has greater clinical utility in the differential diagnosis of peroxisomal disorders based on VLCFA pattern. Plasmalogens will add additional value in differentiating RCDP and Refsum disease.