Dr Marika Kaakinen
Lecturer in Statistical Multi-omics
I hold an MSc degree in statistics and a PhD in genetic and life-course epidemiology from the University of Oulu, Finland. Before joining the University of Surrey in April 2019, I worked as a Marie Curie Fellow, followed by a post as a Research Associate at Imperial College London, UK.
I develop and apply statistical analysis methods for genomic/omics research of complex human traits, including type 2 diabetes and psychiatric traits. I work with various types of omics data, including metabolomics, proteomics, gut microbiome and whole-genome sequencing data. I have developed/contributed to the following software tools: MARV and SCOPA. I have also contributed to numerous GWAS within several consortia, including DIAGRAM (DIAbetes Genetics Replication And Meta-analysis), MAGIC (Meta-Analyses of Glucose and Insulin related traits), ENGAGE (European Network of Genomic and Genetic Epidemiology), EGG (Early Growth Genetics) and SSGAC (Social Science Genetic Association Consortium).
This event will feature two talks:
Multi-phenotype methodology to improve discovery and inference of the genetic architecture of complex human traits
Speaker: Dr Marika Kaakinen, University of Surrey
Genome-wide association studies (GWAS) have been successful in discovering genetic variants associated with thousands of human traits. However, such an approach applied in the traditional sense, i.e. one trait at a time, ignores the correlation between the studied traits. Moreover, correlated traits often share underlying pathways and genetic variants within the pathways. Multi-phenotype GWAS addresses this issue by jointly modelling multiple traits at a time. However, expanding the analysis to high-dimensional data often leads to more sever missing data issues. In this talk I will describe the methods I have developed and contributed to through examples of application to metabolic and psychiatric traits. I will also discuss the challenges in using such approaches, including missing phenotype imputation for high-dimensional data.
Multi-omics for prediction of health and disease, and the role for gut microbiome
Speaker: Dr Ayse Demirkan, University of Surrey
The investigation of the growing population human genetics and lifestyle is in high demand. It is clear now that the more information we obtain, the clearer picture about chronic illnesses predisposition and progression stages we might get. However, the relationship between genetic, environmental, or lifestyle factors is still being investigated. Microbiomics, together with other omics type research is a rapidly developing field where all microorganisms of a given community are studied together. Among the mostly studied, the human microbiome is most often understood as the bacterial community. Under the influence of various factors, the composition of the human microbiota associates and influences diseases. In the section of statistical multi-omics we currently focus on two different projects on gut microbiome, one to elucidate risk factors for Parkinson Disease, and one to describe the associations with markers of health and lifestyle in the Atlas Biomed Cohort.