Congratulations Dr. Nuriye Özlem Özcan Şimşek!

Nuriye Özlem Özcan Şimşek has successfully defended his PhD thesis

Genomic Data Analysis Using Machine Learning Methods For Disease and Disease-Gene Prediction  

Abstract  

Genomic diseases arise due to certain mutations or combinations of mutations in the DNA. This combination can be different for each patient and the effect of each mutation is different for the disease. In this study, we are focussing on the genomic causes of diseases. We defined two research problems. One is to detect the disease from the genetic code, represented as list of mutations. The second is to detect disease-gene associations.

In this thesis, we focussed on cancer and proposed three frameworks. In one approach, the list of mutations was modelled as a document and the mutations were modelled as words in the document. Based on this assumption, we proposed representation models for these patient mutation documents and used them for disease prediction. For another approach, we modelled a novel heterogeneous graph environment in which patients and genes/proteins are the nodes and mutations define the edges. Both approaches to the problem resulted in significantly better classification performances with the selected algorithms, demonstrating the success of our novel designs for the input mutations. The parameters of the two classification frameworks were analysed and a list of the most effective genes for disease prediction was generated for each of these systems. These genes were found to be studied as causal or target genes in the cancer literature. In addition, this list of effective genes was transferred to the gene expression domain as a gene selection algorithm and was found to increase the rate of true prediction of disease. The proposed systems were tested on cancer data but can be easily adapted to other genomic diseases.