Help Privacy Policy Disclaimer
  Advanced SearchBrowse





xploring Genetic Interactions: from Tools Development with Massive Parallelization on GPGPU to Multi-Phenotype Studies on Dyslexia


Jiang,  Beibei
RG Statistical Genetics, Max Planck Institute of Psychiatry, Max Planck Society;

External Resource
No external resources are shared
Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)

Beibei Jiang.pdf
(Any fulltext), 10MB

Supplementary Material (public)
There is no public supplementary material available

Jiang, B. (2019). xploring Genetic Interactions: from Tools Development with Massive Parallelization on GPGPU to Multi-Phenotype Studies on Dyslexia. PhD Thesis, Technische Universität München, München.

Cite as: https://hdl.handle.net/21.11116/0000-000B-40B3-2
Over a decade, genome-wide association studies (GWASs) have provided insightful information into the genetic architecture of complex traits. However, the variants found by GWASs explain just a small portion of heritability. Meanwhile, as large scale GWASs and meta-analyses of multiple phenotypes are becoming increasingly common, there is a need to develop computationally efficient models/tools for multi-locus studies and multi-phenotype studies. Thus, we were motivated to focus on the development of tools serving for epistatic studies and to seek for analysis strategy jointly analyzed multiple phenotypes. By exploiting the technical and methodological progress, we developed three R packages. SimPhe was built based on the Cockerham epistasis model to simulate (multiple correlated) phenotype(s) with epistatic effects. Another two packages, episcan and gpuEpiScan, simplified the calculation of EPIBALSTER and epiHSIC and were implemented with high performance, especially the package based on Graphics Processing Unit (GPU). The two packages can be employed by epistasis detection in both case-control studies and quantitative trait studies. Our packages might help drive down costs of computation and increase innovation in epistatic studies. Moreover, we explored the gene-gene interactions on developmental dyslexia, which is mainly characterized by reading problems in children. Multivariate meta-analysis was performed on genome-wide interaction study (GWIS) for reading-related phenotypes in the dyslexia dataset, which contains nine cohorts from different locations. We identified one genome-wide significant epistasis, rs1442415 and rs8013684, associated with word reading, as well as suggestive genetic interactions which might affect reading abilities. Except for rs1442415, which has been reported to influence educational attainment, the genetic variants involved in the suggestive interactions have shown associations with psychiatric disorders in previous GWASs, particularly with bipolar disorder. Our findings suggest making efforts to investigate not just the genetic interactions but also multiple correlated psychiatric disorders.