Functional annotations improve the predictive score of human disease-related mutations in proteins

Abstract

Single nucleotide polymorphisms (SNPs) are the simplest and most frequent form of human DNA variation, also valuable as genetic markers of disease susceptibility. The most investigated SNPs are missense mutations resulting in residue substitutions in the protein. Here we propose SNPs&GO, an accurate method that, starting from a protein sequence, can predict whether a mutation is disease related or not by exploiting the protein functional annotation. The scoring efficiency of SNPs&GO is as high as 82%, with a Matthews correlation coefficient equal to 0.63 over a wide set of annotated nonsynonymous mutations in proteins, including 16,330 disease-related and 17,432 neutral polymorphisms. SNPs&GO collects in unique framework information derived from protein sequence, evolutionary information, and function as encoded in the Gene Ontology terms, and outperforms other available predictive methods.

A compound heterozygous mutation in DPAGT1 results in a congenital disorder of glycosylation with a relatively mild phenotype

“…These mutations were highly conserved as indicated by a high phyloP score (conservation score) of 4.9 and 4.8 (Table 2) and the conservation of the amino acids at positions p.Ile29 and p.Leu168 down to the Fruitfly (Figure 3b). We used three bioinformatics based prediction programs such as Polymorphism Phenotyping2 (polyPhen2), 16 Sorting Intolerant from Tolerant (SIFT), 17 and SNPs&GO, 18 to predict the possible causative nature of the identified mutations. The variant c.503T4C was predicted to be likely damaging/deleterious by all three prediction programs, and the variant c.85A4T was predicted to be likely damaging by SIFT and SNPs&GO, while benign by polyPhen2.…”

PDGFRA-mutant syndrome

“…28 P653L PDGFRA can affect protein function, as predicted by four different computational tools. [19][20][21][22] PDGFRA immunoreactivity, variable in inflammatory fibroid polyps and fibrous tumors, and evident in GISTs, was faintly + /negative in non-tumoral tissue (Figures 2e and f, 3g and i, and 4e). Although PDGFRA immunohistochemistry can be problematic, 29 these consistently observed differences in PDGFRA immunoreactivity probably conceal additional events of PDGFRA tumorigenesis in relation to the germline PDGFRA mutation in background.…”

“…The P653L PDGFRA missense mutation was predicted with four different computational tools: 'SIFT', 19 'PolyPhen-2', 20 'SNPs&GO', 21 and PROVEAN. 22 Further details are reported in Supplementary Materials and Methods.…”

Macular corneal dystrophy and associated corneal thinning

“…Potential pathogenicity of missense variants was evaluated using PolyPhen2, 11 MutPred, 12 and SNPs&GO algorithms. 13 Corneal buttons obtained after right penetrating keratoplasties of probands 2, 3, and 4 were dissected, snap frozen in liquid nitrogen, and embedded in Optimal Cutting Temperature Compound. Tissue was cryosectioned at a thickness of 7 mm and stained with haematoxylin-eosin and Alcian blue for morphological assessment by light microscopy.…”