Optimization and Applications of Large-Scale Biomedical Event Networks

The overwhelming amount and unprecedented speed of publication in the biomedical domain make it difficult for life science researchers to acquire and maintain a broad view of the field and gather all information that would be relevant for their research. As a response to this problem, the BioNLP (Biomedical Natural Language Processing) community of researches has emerged and strives to assist life science researchers by developing modern natural language processing (NLP), information extraction (IE) and information retrieval (IR) methods that can be applied at large-scale, to scan the whole publicly available biomedical literature and extract and aggregate the information found within, while automatically normalizing the variability of natural language statements. Among different tasks, biomedical event extraction has received much attention within BioNLP community recently. Biomedical event extraction constitutes the identification of biological processes and interactions described in biomedical literature, and their representation as a set of recursive event structures. The 2009–2013 series of BioNLP Shared Tasks on Event Extraction have given raise to a number of event extraction systems, several of which have been applied at a large scale (the full set of PubMed abstracts and PubMed Central Open Access full text articles), leading to creation of massive biomedical event databases, each of which containing millions of events. Sinece top-ranking event extraction systems are based on machine-learning approach and are trained on the narrow-domain, carefully selected Shared Task training data, their performance drops when being faced with the topically highly varied PubMed and PubMed Central documents. Specifically, false-positive predictions by these systems lead to generation of incorrect biomolecular events which are spotted by the end-users. This thesis proposes a novel post-processing approach, utilizing a combination of supervised and unsupervised learning techniques, that can automatically identify and filter out a considerable proportion of incorrect events from large-scale event databases, thus increasing the general credibility of those databases. The second part of this thesis is dedicated to a system we developed for hypothesis generation from large-scale event databases, which is able to discover novel biomolecular interactions among genes/gene-products. We cast the hypothesis generation problem as a supervised network topology prediction, i.e predicting new edges in the network, as well as types and directions for these edges, utilizing a set of features that can be extracted from large biomedical event networks. Routine machine learning evaluation results, as well as manual evaluation results suggest that the problem is indeed learnable. This work won the Best Paper Award in The 5th International Symposium on Languages in Biology and Medicine (LBM 2013).

Glycosylation and Glycodiversification in Polyketide Antibiotics: Unraveling Biosynthetic Steps in Nogalamycin Formation

Soil-dwelling Streptomyces bacteria are known for their ability to produce biologically active compounds such as antimicrobial, immunosuppressant, antifungal and anticancer drugs. S. nogalater is the producer of nogalamycin, a potential anticancer drug exhibiting high cytotoxicity and activity against human topoisomerases I and II. Nogalamycin is an anthracycline polyketide comprising a four-ring aromatic backbone,a neutral deoxy sugar at C7, and an amino sugar attached via an O–C bond at C1 and a C–C bond between C2 and C5´´. This kind of attachment of the amino sugar is unusual thus making the structure of the compound highly interesting. The sugar is also associated with the biological activity of nogalamycin, as it facilitates binding to DNA. Furthermore, the sugar moieties of anthracyclines are often crucial for their biological activity. Together the interesting attachment of the amino sugar and the general reliance of polyketides on the sugar moieties for bioactivity have made the study of the biosynthesis of nogalamycin attractive. The sugar moieties are typically attached by glycosyltransferases, which use two substrates: the donor and the acceptor. The literature review of the thesis is focused on the glycosylation of polyketides and the possibilities to alter their glycosylation patterns. My own thesis work revolves around the biosynthesis of nogalamycin. We have elucidated the individual steps that lead to its rather unique structure. We reconstructed the whole biosynthetic pathway in the heterologous host S. albus using a cosmid and a plasmid. In the process, we were able to isolate new compounds when the cosmid, which contains the majority of the nogalamycin gene cluster, was expressed alone in the heterologous host. The new compounds included true intermediates of the pathway as well as metabolites, which were most likely altered by the endogenous enzymes of the host. The biological activity of the most interesting new products was tested against human topoisomerases I and II, and they were found to exhibit such activities. The heterologous expression system facilitated the generation of mutants with inactivated biosynthetic genes. In that process, we were able to identify the functions of the glycosyltransferases SnogE and SnogD, solve the structure of SnogD, discover a novel C1-hydroxylase system comprising SnoaW and SnoaL2, and establish that the two homologous non-heme α-ketoglutarate and Fe2+ dependent enzymes SnoK and SnoN catalyze atypical reactions on the pathway. We demonstrated that SnoK was responsible for the formation of the additional C–C bond, whereas SnoN is an epimerase. A combination of in vivo and in vitro techniques was utilized to unravel the details of these enzymes. Protein crystallography gave us an important means to understand the mechanisms. Furthermore, the solved structures serve as platforms for future rational design of the enzymes.