GPCards: An integrated database of genotype-phenotype correlations in human genetic diseases

Abstract

Genotype-phenotype correlations are the basis of precision medicine of human genetic diseases. However, it remains a challenge for clinicians and researchers to conveniently access detailed individual-level clinical phenotypic features of patients with various genetic variants. To address this urgent need, we manually searched for genetic studies in PubMed and catalogued 8,309 genetic variants in 1,288 genes from 17,738 patients with detailed clinical phenotypic features from 1,855 publications. Based on genotype-phenotype correlations in this dataset, we developed an user-friendly online database called GPCards (http://genemed.tech/gpcards/), which not only provided the association between genetic diseases and disease genes, but also the prevalence of various clinical phenotypes related to disease genes and the patient-level mapping between these clinical phenotypes and genetic variants. To accelerate the interpretation of genetic variants, we integrated 62 well-known variant-level and gene-level genomic data sources, including functional predictions, allele frequencies in different populations, and disease-related information. Furthermore, GPCards enables automatic analyses of users' own genetic data, comprehensive annotation, prioritization of candidate functional variants, and identification of genotype-phenotype correlations using custom parameters. In conclusion, GPCards is expected to accelerate the interpretation of genotype-phenotype correlations, subtype classification, and candidate gene prioritisation in human genetic diseases.

Keywords: GPCards; Genotype; Phenotype; Variant.

Fig. 1

A general workflow of GPCards. Data collection and quality control information were showed in green box; Variants annotation and integration flow chart was listed in yellow box; and database construction and interface were exhibited in red box. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Summary of catalogued genotype–phenotype correlation data. (A) The distribution of disease-associated genes with different number of genetic variants. (B) The distribution of studies with different number of clinical phenotypes. (C) The distribution of studies with different number of patients. (D) The distribution of patients with different number of clinical phenotypes.

Fig. 3

Snapshot of search modules in GPCards. The quick search bar is set with 11 types of searches prompts as the example of JAG1. The advanced search could be used to conveniently search in batches with nine type of search prompts. The searching results would show PubMed ID, gene symbol, disorder name, number of variants, patients, and phenotypes. “Specify annotation datasets” is a selectable panel with 24 predictive tools, population-specific allele frequencies, and data from established disease- and phenotype-related databases, allowing users to assign annotation information presented in the panel of searching results.

Fig. 4

Snapshot of genotype-phenotype correlations in GPCards. In “Phenotype Summary and Genotype-Phenotype Correlation” panel, the basic information of the searched genes was presented. The frequencies of various clinical phenotypes or symptoms of disease-causing genes is exhibited in the “Phenotype Summary” panel. The detailed individual-level phenotypes and genotypes were present in “Genotype-Phenotype Correlation” panel. Moreover, comprehensive variant-level annotations of each genetic variant were also present in this panel.