Loading...

Prognostic Biomarker Selection for Breast Cancer using Bioinformatics and Deep Learning

Salimi , Adel | 2021

395 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54049 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Sharifi Zarchi, Ali
  7. Abstract:
  8. Triple Negative Breast Cancer (TNBC) is an invasive subtype of breast cancer. Finding prognostic biomarkers is helpful in choosing the appropriate treatment procedure for patients of this cancer. In recent years, the role of microRNAs in various biological processes, including cancer, has been identified, and their accessibility and stability have made these types of molecules an ideal biomarker. In the first phase of this study, with the aim of overcoming the limitations of previous studies, a new bioinformatics protocol has been proposed to investigate the prognostic miRNAs of triple negative breast cancer. First, using survival analysis, 56 prognostic miRNAs which had a significant association with overall survival were selected. Then, by examining their expression changes in tumor, adjacent normal and real normal tissues, expression changes during cancer progression, extraction of target genes and transcription factors affecting these miRNAs and analysis of biological pathways involved, a comprehensive information about these miRNAs and their biological role was provided.Until now, statistical survival methods such as Kaplan Meier and the Cox proportional hazard model have been used many times, but often the simplicity of these models does not have enough power to make accurate predictions. Deep neural network models, despite having more accurate predictions, have the problem of requiring high dimensional input, which makes their laboratory and clinical use difficult. In the second phase of this study, by developing deep learning methods in survival analysis and adding feature selection to them, it is possible to use these models with limited features and acceptable accuracy to predict patient survival. Compared to previous feature selection methods in survival analysis, the proposed models were able to show a higher concordance index of 0.726 and 0.712 on miRNA and gene data of the TCGA breast cancer dataset, respectively
  9. Keywords:
  10. Deep Learning ; Feature Selection ; Survival Prediction ; Duration Analysis ; Breast Cancer ; Triple Negative Breast Cancer (TNBC) ; Prognostic Biomarker ; Bioinformatics

 Digital Object List

 Bookmark

No TOC