Cystic Fibrosis as a model use case for implementing cell based disease models in systems medicine
by Liza Vinhoven
Date of Examination:2022-11-10
Date of issue:2022-11-22
Advisor:Dr. Manuel Nietert
Referee:Dr. Manuel Nietert
Referee:Prof. Dr. Wolfram-Hubertus Zimmermann
Files in this item
Name:Thesis_LizaVinhoven_digital.pdf
Size:30.8Mb
Format:PDF
Abstract
English
In the last two decades, tremendous progress has been made in producing large amounts of biological and biomedical data in a time- and cost-effective manner. Along with the increasing amount of data has come the requirement for methods to analyze and interpret it effectively. This lead to a multitude of computational methods being developed, often rather sophisticated and highly specific to the datatype and -source. However, in order to utilize the wealth of data to its full potential in systems medicine, it is essential to bring the different data sources together and use approaches, that can easily be applied and adapted to different use cases. Therefore, using the example of Cystic Fibrosis, this thesis focuses on applying generic systems medicine methods to integrate different kinds of data and thereby create a holistic overview of the disease and gain new insights. Cystic Fibrosis is one of the most common genetic diseases prevalent among the white European population. Its vast range of geno- and phenotypes makes the development of therapeutics especially challenging. During the last years, different small-molecule therapeutics have been developed that amplify CFTR function, but they are not effective for all patients. The latest research efforts, therefore, focus on developing combination therapies to target multiple defects at once. To provide an overview of already tested compounds, I contributed to establishing the publicly available database CandActCFTR, where substances are listed and categorized according to their interaction with CFTR. It becomes apparent that for the majority of compounds it is unknown whether they affect CFTR directly or indirectly. To elucidate the mechanism of action for promising candidate substances and be able to predict possible synergistic effects of substance combinations, I created a systems medicine disease map of the CFTR biogenesis, function and interactions. In order to support the manual curation and upkeep of disease maps, a tool was developed to integrate text mining approaches into the disease map curation. The tool allows the user to iterate through the text mined interactions to validate the results, thereby bringing together the speed of text mining and the accuracy of scientific expert knowledge. To bring together the chemical knowledge from the database and the biological pathways from the disease map, an interlinking tool was developed, to interactively and computationally map compounds to their respective targets based on publically available interaction data. This data, however, still leaves the mechanism of action for the majority of the active compounds unexplained. Therefore, in order to suggest possible modes of action for all active compounds in the database, I used two complementary in silico target identification approaches, namely target-based molecular docking and ligand-based similarity searches. I thereby identified possible targets for all active compounds, which will help to understand which compound classes affect CFTR at which stage of its life cycle and which compounds can be combined to alleviate different defects in its biogenesis. All parts of the project can be seen individually as stand-alone resources and be combined to yield new findings. Since the approaches are generic and easily adaptable, they can also be applied to other disease besides Cystic Fibrosis in a similar manner, to give a holistic, systems medicine approach to answer research questions relevant to them.
Keywords: Systems Medicine; Cystic Fibrosis; Bioinformatics; Systems Biology; Computational Biology