Identification of novel biomarkers and candidate genes associated to lipid traits : improving the lipid metabolism knowledge base

• Main Author: Correia, Marta
• Publication Date: 2022
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10451/54984
• topic: hipercolesterolemia familiar; biomarcadores; perfil lipídico estendido; métodos baseados em machine learning; base de dados de conhecimento lipídico; familial hypercholesterolaemia; biomarkers; extended lipid profile; machine-learning based methods; lipid knowledge base; Domínio/Área Científica::Ciências Naturais::Ciências Biológicas

FH, the most common monogenic dyslipidaemia, is characterised by increased circulating LDL-C levels leading to premature cardiovascular disease when undiagnosed or untreated. Current guidelines support genetic testing in patients fulfilling clinical diagnostic criteria and cascade screening of their family members. However, about half of clinical FH patients do not present pathogenic variants in the known disease genes (LDLR, APOB, PCSK9), and these most likely suffer from polygenic hypercholesterolaemia, which translates into a relatively low yield of genetic screening programs. This project aimed to identify new biomarkers able to improve the distinction between monogenic and polygenic profiles. Using a machine-learning approach in a paediatric dataset, tested for disease causative genes and investigated with an extended lipid profile, we developed new models that classify FH patients with higher specificity than currently used methods. The best performing models incorporated parameters absent from the common FH clinical criteria, which rely only on TC and LDL-C. A hierarchical clustering analysis of the same dataset showed that the study population can be clearly divided in three groups of dyslipidaemic individuals, showing the complexity of the dyslipidaemic biological context and the need of an integrative and multidisciplinary approach for biomarker selection. Both clustering and modelling analysis have revealed that the extended lipid profile contains important biomarkers. The exploration of lipid metabolic pathways associated with the identified biomarkers allowed us to identify a set of related genes. Using additional information from public databases, including gene expression data, associated GWAS and GO terms, we defined a universe of lipid-related genes and molecular interactions relevant for the dyslipidaemic context and future genetic studies. All this information was used to establish a new lipid knowledge base available online. The obtained results can be applied to improve the yield of genetic screening programs and decrease the associated costs, and also provide novel contributions to our understanding of dyslipidaemias.