K. Lakiotaki (1), E. Kartsaki (1), A. Kanterakis (1), T. Katsila (2), G. P. Patrinos (2), G. Potamias (1)
(1) Institute of Computer Science, Foundation for Research & Technology – Hellas, Heraklion, Crete (2) Department of Pharmacy, University of Patras, Hellas
Objectives: To develop an integrated electronic Pharmacogenomics (PGx) web service that provides personalized genotype-to-phenotype translation services, linked to drug recommendations.
Methods: Genomic information repositories are mainly developed and maintained by stable governmental funded efforts, along with data sharing services (e.g., NCBI’s dbSNP, dbGaP, dbVar and ClinVar). In contrast, the curation of genes related to the absorption, distribution, metabolism, and excretion – toxicity (ADMET) of drugs are usually handled by locus-specific databases such as, the Human Cytochrome P450 (CYP) Allele Nomenclature (www.cypalleles.ki.se), or databases focusing on transporters, receptors, and kinases [1]. Other online resources offer documented frequencies of pathogenic genetic variations that lead to inherited disorders in various populations worldwide, e.g., FINDbase (www.findbase.org) [2]. We developed automatic tools that combine data from heterogeneous sources (PharmGKB, dbSNP, Ensemble) and link them to available, up to date, pharmacogenomics information [3]. These tools are integrated into a single web portal, where users can query for combinations of genes, drugs and alleles, and browse related clinical guidelines. Moreover, users may upload genotypes in Variant Call Format (VCF) and receive personalized drug recommendations, when available.
Results: To demontrate the integrated electronic Pharmacogenomics web service, we explored genome data from phase3 1000 Genomes Project (1kG). Statistical analysis shows a wide population differentiation in ADMET variants among 1kG populations. In general, individuals of African ancestry exhibit greater pharmacogenomics profile differentiation in most ADMET genes, which can be attributed to increased genetic heterogeneity of African population.
Conclusions: We developed an integrated electronic PGx assistant that offers personalized diagnostics based on genomic evidence. The novelty of its approach rests in its ability to infer an individual’s phenotype (metabolizer status) based on a corresponding genotype profile and PGx data that are open and free. Based on this connection, the system acts as a “one stop shop” web portal for clinicians – by supporting them in making informed decisions, and for researchers – by providing a single place with information to understand, document and assess individuals’ differences in drug efficacy.
References:
[1] S. Sim, R. Altman, and M. Ingelman-Sundberg, “Databases in the area of pharmacogenetics”, Human mutation, 32, 5, (2011).
[2] P. Papadopoulos, E. Viennas, V. Gkantouna, C. Pavlidis, M. Bartsakoulia, Z.-M. Ioannou, I. Ratbi, A. Sefiani, J. Tsaknakis, K. Poulas, G. Tzimas, and G. P. Patrinos, “Developments in FINDbase worldwide database for clinically relevant genomic variation allele frequencies”, Nucleic Acids Research, 42, Database issue, (2014).
[3] K. Lakiotaki, G. P. Patrinos, and G. Potamias, “Information technology meets pharmacogenomics: Design specifications of an integrated personalized pharmacogenomics information system” in IEEE-EMBS International Conferences on Biomedical and Health Informatics, (2014)
Reference: PAGE 24 (2015) Abstr 3652 [www.page-meeting.org/?abstract=3652]
Poster: Drug/Disease modeling - Other topics