Machine-learning (ML) techniques have been widely applied to solve different problems in biology. However, biological data are large and complex, which often result in extremely intricate ML models. Frequently, these models may have a poor performance or may be computationally unfeasible. This study presents a set of novel computational methods and focuses on the application of genetic algorithms (GAs) for the simplification and optimization of ML models and their applications to biological problems. The dissertation addresses the following three challenges. The first is to develop a generalizable classification methodology able to systematically derive competitive models despite the complexity and nature of the data. Although several algorithms for the induction of classification models have been proposed, the algorithms are data dependent. Consequently, we developed OmniGA, a novel and generalizable framework that uses different classification models in a treeXlike decision structure, along with a parallel GA for the optimization of the OmniGA structure. Results show that OmniGA consistently outperformed existing commonly used classification models. The second challenge is the prediction of translation initiation sites (TIS) in plants genomic DNA. We performed a statistical analysis of the genomic DNA and proposed a new set of discriminant features for this problem. We developed a wrapper method based on GAs for selecting an optimal feature subset, which, in conjunction with a classification model, produced the most accurate framework for the recognition of TIS in plants. Finally, results demonstrate that despite the evolutionary distance between different plants, our approach successfully identified conserved genomic elements that may serve as the starting point for the development of a generic model for prediction of TIS in eukaryotic organisms. Finally, the third challenge is the accurate prediction of polyadenylation signals in human genomic DNA. To achieve this, we analyzed genomic DNA sequences for the 12 most frequent polyadenylation signal variants and proposed a new set of features that may contribute to the understanding of the polyadenylation process. We derived Omni-PolyA, a model, and tool based on OmniGA for the prediction of the polyadenylation signals. Results show that Omni-PolyA significantly reduced the average classification error rate compared to the state-of-the-art results.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第21654号 / 医博第4460号 / 新制||医||1035(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 朝長 啓造, 教授 妹尾 浩, 教授 萩原 正敏 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

No description available.

No description available.

No description available.

No description available.

Translation initiation in eukaryotes is a multistep process requiring the orchestrated interaction of several eukaryotic initiation factors (eIFs) together with the small ribosomal subunit to locate the mRNA's translational start and to properly decode the genetic message that it carries. The largest of these factors, eIF3, forms the scaffold for other initiation factors to promote their spatially coordinated placement on the ribosomal surface. It is our long-standing pursuit to map the 40S-binding site of the yeast multisubunit eIF3 and here we present three new mutual interactions between these two macromolecules (i) The C-terminal region of the eIF3c/NIP1 subunit is comprised of the conserved bipartite PCI domain and we show that a short C-terminal truncation and two clustered mutations directly disturbing the PCI domain produce lethal or slow growth phenotypes and significantly reduce amounts of 40S-bound eIF3 in vivo. The extreme C-terminus directly interacts with small subunit ribosomal protein RACK1/ASC1, which is a part of the 40S head, and, consistently, deletion of ASC1 impairs eIF3 association with ribosomes. The PCI domain per se shows strong but unspecific binding to RNA, for the first time implicating this protein fold in protein-RNA interactions. We conclude that the c/NIP1...

Translation is the process by which proteins are synthesized from the instructions in the genetic code. Translation is mediated by the ribosome, a large ribonucleoprotein complex, in concert with messenger RNA (mRNA), transfer RNA (tRNA), and a variety of proteins. The canonical mechanism of translation, introduced in Part I of my thesis, is divided into four distinct phases: initiation, elongation, termination, and recycling. Under unusual circumstances, each phase of translation can also proceed via a number of noncanonical mechanisms, many of which are vitally important for cellular growth or viral infectivity. My thesis describes structural insights into two such noncanonical mechanisms. The aim of the first project, described in Part II, was to structurally characterize a noncanonical mechanism of translational termination in bacteria. In the absence of a stop codon, ribosomes arrest at the 3′ end of an mRNA and are unable to terminate. In bacteria, the primary mechanism for rescuing such nonstop complexes is known as trans-translation. In the absence of a functional trans-translation system, however, the small protein ArfA recognizes the empty mRNA channel and recruits the release factor RF2 to the ribosome, enabling termination to occur. Using single-particle electron cryomicroscopy (cryo-EM), I obtained four high-resolution structures of nonstop complexes that reveal the mechanism of ArfA-mediated ribosome rescue and have wider implications for understanding canonical termination in bacteria. The aim of the second project, described in Part III, was to gain structural insights into a noncanonical mechanism of translational initiation in eukaryotes known as internal ribosome entry. Instead of a 5′ cap, many viruses contain intricately structured, cis-acting internal-ribosome-entry sites (IRESs) within their genomes that direct end-independent initiation. The IRES of hepatitis-C virus (HCV), for example, interacts directly with the mammalian ribosome and functionally replaces many of the canonical initiation factors. However, the mechanism by which the HCV IRES coordinates assembly of an initiation complex and progresses through the initiation phase remains poorly understood. I developed a method for purifying native ribosomal complexes from cell lysate that enabled me to obtain multiple cryo-EM maps of the HCV IRES in complex with the 80S ribosome, including a previously unseen conformation of the IRES induced by rotation of the ribosomal small subunit, and to make progress towards capturing earlier steps in the initiation pathway.

Protein synthesis is a complex process performed by sophisticated cellular particles known as ribosomes. Although RNA constitutes the major structural and functional component, ribosomes from all kingdoms contain an extensive array of proteins with largely undefined functional roles. The work presented in this thesis addresses ribosomal complexity using mutants of Salmonella typhimurium to examine the physiological effects of ribosomal protein (r-protein) removal and orthologous replacement on bacterial fitness and ribosome function. The results of paper I demonstrate that removal of small subunit protein S20 conferred two independent translation initiation defects: (i) a significant reduction in the rate and extent of mRNA binding and (ii) a drastic decrease in the yield of 70S complexes caused by an impairment in subunit association. The topographical location of S20 in mature 30S subunits suggests that these perturbations are the result of improper orientation of helix 44 of the 16S rRNA when S20 is absent. In paper II we show that the major functional impairment associated with loss of large subunit protein L1 manifested as an increase in free ribosomal subunits at the expense of translationally active 70S particles. Furthermore, the formation of free ribosomal subunits was imbalanced suggesting that L1 is required to suppress degradation or promote formation of 30S subunits. Compensatory evolution revealed that mutations in other large subunit proteins mitigate the cost of L1 removal, in one case seemingly via an increase in 70S complex formation. As shown in paper III, the large fitness costs associated with complete removal of r-proteins is in contrast to the generally mild costs of orthologous protein replacement, even in the absence of a high degree of homology to the native protein. This clearly demonstrates the robustness and plasticity of the ribosome and protein synthesis in general and it also implies that functional constraints are highly conserved between these proteins. The findings of paper III also allowed us to examine the barriers that constrain horizontal gene transfer and we find that increased gene dosage of the sub-optimal heterologous protein may be an initial response to stabilize deleterious transfer events. Overall the results highlight the requirement of r-proteins for the maintenance of ribosomal structural integrity.

Title from second page of PDF document. Includes bibliographical references (p. Xx-Xx).