PPML (PKU Phenotype prediction by Machine Learning) is a general framework to predict the PKU phenotype based PAH genotype. Phenylketonuria (PKU, OMIM# 261600) is a common autosomal recessive inherited metabolic disease with an inborn error of phenylalanine (Phe) metabolism, which is caused by pathogenetic variants in the phenylalanine hydroxylase (PAH) gene. PPML training case database for classified the PKU phenotype into three categories cPKU, mPKU and HPA. The features are extracted from the information of nucleotide mutations and amino acid change information, as well as the property of the allelic mutation linkage graph for training with machine learning classification models. PPML provides a powerful analytical tool for clinical analysis and inference of the identification of PAH mutations causing PKU with compound heterozygotes as well as heterozygotes.

Citation: Fang, Yang, et al. "Allelic phenotype prediction of phenylketonuria based on the machine learning method." Human Genomics 17.1 (2023): 1-9. PDF