Lysosomal storage diseases (LSDs) are a group of genetic disorders, resulting from deficiencies of lysosomal enzyme. Genotype-phenotype correlation is essential for timely and proper treatment allocation. Recently, by integrating prediction outcomes of 7 bioinformatics tools, we developed a SAAMP algorithm to predict the impact of individual amino-acid substitution. To optimize this approach, we evaluated the performance of these bioinformatics tools in a broad array of genes. PolyPhen and PROVEAN had the best performances, while SNP&GOs, PANTHER and I-Mutant had the worst performances. Therefore, SAAMP 2.0 was developed by excluding 3 tools with worst performance, yielding a sensitivity of 94% and a specificity of 90%. To generalize the guideline to proteins without known structures, we built the three-dimensional model of iduronate-2-sulfatase by homology modeling. Further, we investigated the phenotype severity of known disease-causing mutations of the GLB1 gene, which lead to 2 LSDs (GM1 gangliosidosis and Morquio disease type B). Based on the previous literature and structural analysis, we associated these mutations with disease subtypes and proposed a theory to explain the complicated genotype-phenotype correlation. Collectively, an updated guideline for phenotype prediction with SAAMP 2.0 was proposed, which will provide essential information for early diagnosis and proper treatment allocation, and they may be generalized to many monogenic diseases.
Keywords: GM1 gangliosidosis; Morquio syndrome; genotype-phenotype correlation; homology modeling; in silico; lysosomal storage disease.
© 2018 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.