Loading...

Analyzing Microarray Data Via Learning DNA Cross Hybridization

Hassani Bidgoli, Mansoor | 2017

1853 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49428 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Motahari, Abolfazl; Rabiee, Hamid Reza
  7. Abstract:
  8. Gene expression microarrays include thousands of probes spotted on their surface to measure the expression level of a set of genes. Identifying the amount of a transcript level by hybridization, each probe is complementary to a fragment of a specific gene transcripts. Although probes are designed to avoid crosshybridization to non-specific transcripts, occurrences of cross-hybridizations is inevitable due to massive probes that are spotted on microarrays. The main question is whether these non-specific cross-hybridization have significant effect on the downstream analysis of gene expression microarray datasets. This thesis aims at answering to this question by considering datasets from RNA-Seq technology as reference points. This is due to the fact that RNA-Seq datasets provides much more accurate measurements of gene expressions. To this end, we first propose a novel method to normalize datasets from microarray and RNASeq platforms simultaneously. The normalization brings the datasets to the same level such that most of the gene expressions becomes relatively the same in the two platforms. Then looking at the expression levels of non-matched probes it is hypothesized that they are coming from cross-hybridization with non-specific transcripts. By performing statistical analysis on several hypothetical models, it is shown that cross-hybridization is not the dominant effect on observing different levels of gene expressions in the two platforms
  9. Keywords:
  10. Microarray ; Probe ; Microarray Data Analysis ; Linkage Learning ; Specific Hybridization ; RNA Sequencing ; Cross Hybridization ; Specific Hybridization ; RNA Sequencing

 Digital Object List

 Bookmark

...see more