Analysis of genomic translation using a communications theory approach

The increase in genetic data during the last years has prompted the efforts to use methods and techniques from the engineering fields for their interpretation. These fields include information theory, communications, coding theory, signal processing, machine learning, and various statistical methods. This thesis discusses the role and contributions of communications theory in the field of genetic data analysis. Novel use of techniques and principles from the latter field are proposed for examining and analyzing the genomic structure including coding and non-coding regions. The developed analyses allow testing different biological aspects related to the process of gene translation and determining whether regions of a specified genome are protein-producing sequences. This analysis promotes new interdisciplinary collaborations in research and education, integrating biomedical engineering, electrical engineering and life sciences. The knowledge gain can help address fundamentally important issues that cannot be explored systematically and quantitatively by experimentation alone. Moreover, it can allow savings in laboratory resources and time-consuming laboratory experimentations and leads to better understanding of the complex genetic processes. This thesis deals with modeling the process of translation in gene expression using a communications theory approach. It first investigates a variable length codebook model for the process of translation with an exponentially weighted algorithm to optimize its detection mechanism. A mutational analysis is carried out to test the model and certify its correctness and biological relevance. Different communications theory based metrics are used to quantify the mechanism that the ribosome uses to detect the translational signal. Next, a block code and a convolutional code models for the process of translation are proposed. The two models are based on the same key biological element with differences in model construction, base mapping, and significance. Finally, this thesis proposes two novel methods for regulatory sequence detection. The developed models were tested on different bacterial genomes. The obtained results prove the validity, significance and biological relevance of the models being confirmed with experimental published data. This further verifies the relevance of using communications theory to approach several biological problems and hence encourages the research cooperation between the two communities of communications engineering and molecular biology