ERVs in Bos taurus

Project leader: Chad Harland

Students: Alicia L. Bruzos, Gaston Arocena, Nadezhda Potapova

Description: Recent research has shown Endogenous Retroviruses (ERVs) are active in the genome of Bos Taurus (Cattle) and this activity can result in the creation of new phenotypes. It was also shown that these active ERVs were technically non-functional due to the disruption of their genes but were still able to move, suggesting older ERVs could be reactivated by proteins from an external source. The aim of this project was to look at three populations of cattle and determine if the active or unique ERVs in each population were more closely related within populations or between populations. The assumption was that if events clustered within populations it supports the hypothesis that single ERV are reactivated and then remains active for some time period, while a lack of clustering suggests that ancestral ERVs were reactivated relatively often in any population... (read more)

Presentation during the conference

Genetic risk profiling within and between traits

Project leader: Peter Joshi
Project tutor: Amy Rowlatt
Students: Andrew Bretherick, Ion Frias, Lauren Navrady, Marina Slashcheva, Tim Padvitsky

Description: Estimated heritability of a trait is usually much bigger than the amount of phenotypic variance that can be explained by results of GWAS or genome-wide association meta-analysis (GWAMA) studies. Polygenic risk scores have been previously shown to capture a greater proportion of overall genetic contribution to a specific trait than individual SNPs from GWA(MA)S alone.

In our project we used published genome-wide association summary statistics to determine the genetic risk profiles for participants in the Orkney Complex Disease study and then we evaluated how predictive the scores are for the trait concerned and other related and apparently unrelated traits. We have also investigated p-value criteria for subsetting SNP's from the pool of all GWA(MA)S SNP's to use for polygenic risk score calculation... (read more)

Presentation during the conference

Integration of several omic sources for phenotypic prediction

Project leaders: Mar Rodríguez-Girondo, Jeanine Houwing-Duistermaat
Project tutors: Alexander Kurilschikov, Andrea Gelemanovic
Students: Ivet Gazova, Gloria Kraus, Pauline Pavlovich, Alexander Tkachenko

Description: We have investigated different methods for integrating several omic sources (glycomics and metabolomics) for phenotypic prediction. By using the within omic dataset structure, we aimed for interpretable models which also have good prediction performance. Among others, we investigated group penalization based on network analysis and combination of single-omic predictions. We used simulated and real datasets to illustrate the performance of each method and to compare them with single-source prediction and naïve combination methods based on stacking... (read more)

Presentation during the conference

Functional genomics analysis of disease-associated loci

Project leaders: Dima Alexeev
Project tutors: Dima (junior), Sodbo Sharapov, Lucija Klaric
Students: Natalya Klimenko, Dunja Habulin, Xue Li

Description: In our project we were using summary data from GWAS and eQTL from independent studies. The aim was to distinguish linkage from pleiotropy and causality (pleiotropy and causality were not distinguished in this project)...

(read more)

Presentation during the conference

Please see the projects of RSSSO2015 and the projects of RSSSO2014 if interested how the projects went in 2015 and 2014