UiB Blogg            

Past dissertations


Past dissertations

Pål Vegard Johnsen
Dissertation: 21th January
PhD thesis title:

 
Neann Mathai
Dissertation: 10th December at 10:30
PhD thesis title: “Development, validation and application of in-silico methods to predict the macromolecular targets of small organic compounds”.
 
Summary of thesis:
Medicines are often small chemicals which interact with larger chemicals, usually proteins, in the body to produce a therapeutic effect. Historically, the discovery of medicines has been serendipitous and many of the “low hanging fruit” have already been identified. Discovering new medicines is therefore highly expensive and risky. In addition, a chemical can interact with several proteins in the body. These wider interactions mean that one chemical could be used to treat several ailments, but it can also produce undesirable side effects. As a result, computational methods are increasingly being used in drug discovery, to predict interactions between chemicals and proteins. These predictions help focus research on chemicals that are likely to interact with proteins of interest, or target proteins, while avoiding interactions with other proteins, or off-target proteins.
 
There are a range of approaches to predict the interactions between chemicals and target proteins. This dissertation first examines how different validation strategies could be used to measure a target prediction method’s performance for a chemical of interest, with its structural specificities, compared to the average query chemical in the test data. The dissertation then developed a similarity based and a random forest based target prediction approach and compared their performances. This research found that the similarity-based approach generally performed better under all testing scenarios, while also having a target coverage which was twice as large. Finally, this research applied the similarity based target prediction approach, and an artificial intelligence algorithm, to curate libraries of potent chemicals. The curated libraries consist of chemicals which are more likely to interact with a wide range of proteins, narrowing the search and helping support medicinal research going forward.
 
 

Samaneh Abolpour Mofrad

Dissertation: 26th October 2021 at 13:15
PhD thesis title: “Learning and cognition in brain and machine – Prediction of dementia from longitudinal data and modelling memory networks”
 
Summary of thesis:
Starting in the mid-20th century and throughout their developments, modern neuroscience and artificial intelligence (AI) have provided each other with inspiration, insights, and tools. The degree to which they are intertwined has been in constant flux over the years, but always present. With the enormous resurgence of interest in machine learning over the past decade, led by the much-celebrated successes of artificial neural networks and deep learning, the bond between the two fields seems to be growing stronger.
 
Artificial intelligence and machine learning have always kept an eye on biological intelligence and learning, as these provide our only examples of general intelligence and strong learning capabilities, inspiring the development of their much less capable–albeit improving–counterparts, which are based on computational models. The growing attention to both neuroscience and AI is also leading to growth where they intersect, i.e. in neuroscience-inspired AI and AI-inspired neuroscience, and in the usage of computational AI models within neuroscience and the cognitive sciences.
 
In this context, the present thesis aims to make a modest contribution through our application of machine learning techniques to the study of dementia using data from longitudinal MRI and psychometric testing, and through our proposed models for simulating aspects of the formation of memory networks during learning and memory retrieval. The former is our main contribution and is addressed in studies A, B, and C, while the latter is reflected in studies D and E.
 
Through longitudinal studies, i.e. studies based on the collection of repeated measurements from the same subjects, or experimental units, over time one can observe how measurements develop and discover new relationships between variables. Longitudinal data analysis is a large field of research comprised of a multitude of methods and is widely applicable to e.g. behavioural analysis and medicine. One inherently longitudinal phenomenon of particular interest for the present work is the biological, neurological, and cognitive alteration linked to aging. There is an immense need to develop methods that can indicate the risk of developing aging-related diseases such as dementia, as well as for increasing the understanding that is derived from new computational models for cognitive skills such as memory and learning.
 
 
Dani Beck
Dissertation: 5th October 2021 at 10:15
PhD thesis title: “Cardiometabolic health and the ageing brain: Using brain and body MRI to elucidate the body-brain axis in healthy adults”
 
Summary of thesis:

The health of the body and the brain are intrinsically connected, with the structure and function of the ageing human brain being vulnerable to the effects of poor cardiovascular health. Cardio- vascular and -metabolic risk factors such as smoking, dyslipidemia, high blood pressure, obesity, and markers of inflammation are associated with an increased risk of neurocognitive conditions such as dementia and stroke, in addition to a range of mental disorders and age-related cognitive decline.

While poor cardiometabolic health may negatively impact brain health, strategies that promote cardiometabolic health may conversely halt the onset of age-related pathological changes in the brain. Increasing knowledge about the mechanisms of cardiometabolic risk factors (CMRs) and their association with brain structure and integrity is necessary for the development of treatment strategies that delay ageing-related neurodegeneration.

In the current thesis, we used a combination of cross-sectional and longitudinal datasets to investigate brain and cardiometabolic health. Utilising brain age prediction based on neuroimaging data (Franke, Ziegler, Kloppel, Gaser, & Alzheimer’s Disease Neuroimaging, 2010), we first tested the age prediction accuracy of brain age models based on diffusion magnetic resonance imaging (MRI) metrics, followed by investigating brain age gap (BAG, the difference between the brain-predicted age and chronological age) associations with CMRs, including clinical measures, blood test measures, and measures of adipose tissue distribution from body MRI.

 
Emmanuel Eduardo Moutoussamy at UiB
Dissertation: 16th September 2021 at 10:15
PhD thesis title: “Computational Close-up on the Interactions Between Phospholipases and Choline-containing Lipids “

John Zobolas at NTNU

Trial lecture: 10th August 2021 at 10:15
Topic for trial lecture: “Assets and drawbacks of commonly used computational frameworks in cancer modeling”
 
Dissertation: 10th August 2021 at 13:15
PhD thesis title: “Software implementations allowing new approaches toward data analysis, modeling and curation of biological knowledge for Systems Medicine”
 

Summary of thesis:

Cancer is one of the leading causes of death globally. To combat cancer, scientists want to understand the mechanisms that drive this disease so that effective treatments can be developed. However, regulatory mechanisms of cancer can be very complex, so its analysis has proven to be quite challenging. The field of Systems Medicine has emerged to address this problem, combining expertise from different scientific disciplines such as Biology, Medicine and Computer Sciences. This approach aims to integrate, analyze and interpret data from different resources and produce new knowledge, which is the basis for the development of personalized therapies. The efforts described in this PhD thesis have improved various parts of a systems medicine approach towards cancer therapy development.

We developed software tools that enable a better collection and sharing of biological knowledge described in scientific publications. Our software facilitates the task to search and annotate information about molecules and their interactions and to store it in a format computers can understand. With this knowledge scientists can build networks describing the structure and signaling information of biological systems such as the human cell. These signaling networks can be used to construct computational cancer models, which can explain why signals traveling through these networks get disrupted, leading to disease. We implemented simulation software that automatically builds and optimizes many models to match given cancer signaling data. Our computer models successfully predicted drug combinations that act synergistically in experiments. A new software tool was built to analyze the simulation data and find biological or modeling markers that explain why some drug combinations are beneficial while others are not. Lastly, we proposed a mathematical framework that allows us to optimally set the parameters of our models and make them better match experimental observations and comply with the underlying signaling information.

 

Einar Marius Hjellestad Martinsen

Dissertation: 25th June 2021 at 12:15
PhD thesis title: “The mycobiome in COPD: Descriptive and longitudinal analysis, and participation in research bronchoscopy studies”
 
Eva Lena Fjell Estensmo
Dissertation: 24th June 2021 at 13:15
PhD thesis title: “The diversity and seasonality of the indoor mycobiome”
 
 
Bram Burger at University of Bergen
Dissertation: June 10tht, 2021 10:15, Zoom
PhD thesis title: Statistical considerations for the design and interpretation of proteomics experiments
 
Ivana Mikocziova at University of Oslo
Trial lecture: May 21st, 2021 10:15, Zoom
Topic for trial lecture: Immune responses to acute viral infections
 
Dissertation:  May 21st, 2021 12:15
Title of thesis: “Characterisation of germline immunoglobulin variants from naïve B cell receptor repertoires

 

Summary of thesis:

Antibodies and B cell receptors are crucial components of the immune system.  However, immunoglobulin genes, which encode both antibodies and B cell receptors, are not well explored. This is partly due to high similarities among immunoglobulin genes, large structural variation of the immunoglobulin loci as well as the presence of somatic hypermutation in antigen-experienced B cells.

In this thesis, we used a dataset composed of naïve immunoglobulin repertoires from a cohort of 100 Norwegians. Since naïve B cells have not undergone somatic hypermutation, we were able to infer germline alleles from this dataset. We optimised existing software tools for germline inference and discovered a great amount of structural variation and a large number of previously unreported immunoglobulin V alleles in both heavy and light chain genes. On top of that, we developed an approach for filtering potential sequencing and PCR artefacts.

We also analysed the leader sequences and 5’ untranslated regions (5’UTR) of immunoglobulin genes and discovered even more polymorphisms in these regions. Surprisingly, we found several sequences with identical coding V region that had different 5’ UTRs and/or leader sequences. Our analysis also revealed alternatively spliced transcripts of genes with low usage levels, which raises questions about mechanisms that regulate the expression of immunoglobulin genes.

The results of our work provide a valuable contribution to the efforts to characterise germline immunoglobulin gene variants. This will enable further research into the functional effect of immunoglobulin polymorphisms as well as exploration of possible influence of coding and non-coding immunoglobulin polymorphisms on health and disease.

 
Ying Yao at University of Oslo
Trial lecture: March 24th 2021 at 09.15
Topic for trial lecture:
The role of gut microbiota in health and disease
 
Dissertation:  March 24th 2021 at 11.15
Title of thesis:“High-throughput sequencing of gluten-specific T cells in celiac disease

Summary of thesis:

Celiac disease is a chronic inflammatory disorder resulting from mis-appropriate immune response toward ingested gluten proteins. Gluten-specific CD4+ T cells, as the key drivers for the pathogenesis can be identified by staining with HLA-DQ:gluten tetramers.

In this project, we conducted single cell transcriptome sequencing on CD4+ T cells sampled from peripheral blood of untreated CD patients. The tetramer-specific T cells showed transcriptomic profiles consistent with activated effector memory T cells that share features with Th1 and follicular helper T cells. Compared to non-specific cells, gluten-specific T cells showed differential expression of several genes involved in metabolic processes, including fatty acid metabolism and redox potentials. In addition, we utilized the unique expression of immune receptor of each T and B cell to quantify the impact of index switching on single cell RNA-seq experiments.

We also demonstrated that the state of celiac disease could be inferred by unbiased direct TCR sequencing on lamina propria T cells, which is promising for the ultimate goal of inferring celiac disease state based on TCR sequencing of circulating T cells.

 

Vanessa Carina Bieker

Dissertation:  March 19th 2021
Title of thesis: “Using historical herbarium specimens to elucidate the evolutionary genomics of plant invasion”

Summary of thesis:

The world’s herbarium collections contain a vast number of specimens that were collected up to 400 years ago. Due to recent advances in DNA extraction and sequencing, these specimens are now readily available for use in genomic studies. This allows us to directly sample past diversity and reveal evolutionary and population histories that are difficult or even impossible to infer from modern data alone.

 

Due to globalization, increasing numbers of  species are introduced into locations they could not reach through natural dispersal. Some of these introduced species are able to establish a stable population in the introduced range and eventually become invasive. These species may be able to outcompete native species and thus can drive them to local or even global extinction. Therefore, invasive species are a threat to global biodiversity. Despite this, invasive species are good study systems for evolutionary processes. Each introduction to a new environment can be viewed as a natural experiment that is often running for more generations than can be studied e.g. in experimental evolution studies. Thus, parallel adaptation to similar environments can be studied. In combination with archaeobotanical samples or historic herbarium records, these changes can be observed directly using  genetic evidence. One hypothesis as to why invasive species are successful in the introduced range is the Evolution of Increased Competitive Ability (EICA) hypothesis. It says that due to the release from native enemies, plants are able to allocate resources away from defense mechanisms and towards increased growth and reproduction which makes them better invaders.

 

Ambrosia artemisiifolia (common ragweed) is a very successful invasive annual weed native to North America that was introduced to Europe in the late 19th century. In this thesis, I investigate the genomic basis of this invasion using 297 historic herbarium specimens and 350 contemporary samples from both the native and the invasive range in what is thus far the largest study of whole genomes of a non-model, non-crop plant. I found that the population structure in the native range contains three main genetic clusters and one admixed cluster, and, using this information, I was able to identify the most likely source population for the European invasion. Unlike in the native range, population structure changed drastically over time in Europe. Moreover, I found selection on traits in Europe which are consistent with the EICA hypothesis. In addition, I used the herbarium specimens as well as the contemporary samples in the first study of its kind to look at differences of the metagenomic community between ranges and over time. I found that some taxa are less common in Europe, providing evidence that enemy release might have played a role in the plants’ invasive success. This thesis demonstrates how genomic data from herbarium specimens can be used in a variety of  studies and how they add value to the study of invasion.

 

Martin Wohlwend

Dissertation:  March 19th 2021
Title of thesis: “Exercise and the dark human genome reveal gene-targets in age-related diseases”
Summary of thesis:

Therapeutic efficacies for age-related diseases such as heart failure, diabetes and sarcopenia are inadequate because heart failure is still the main cause of death worldwide, prevalence of diabetes is increasing, and, there are no therapeutic options to combat sarcopenia. Therefore, innovative approaches to contest these age-related diseases are critically needed. This thesis explored exercise-mechanisms and the noncoding genome to identify novel genes implicated in age-related diseases with clear therapeutic potential.

A large body of evidence indicates exercise as an undisputed mediator of cardiac/skeletal muscle health. However, molecular mechanisms underlying the benefits of exercise are not fully understood. Hence, harnessing exercise-mechanisms might present an intriguing strategy to therapeutically deliver some specific benefits of exercise.

An additional treasure trove for gene target discovery opened up when the human genome was sequenced in 2003, shockingly showing that 99% of DNA is noncoding. The noncoding part of DNA is not well studied and has therefore been termed the “dark genome”. The dark side of the human genome partly consists of enhancer elements, which contain the vast majority of genetic variants associated with common diseases. Another intriguing entity comprised in this unknown part of our genome are large transcribed, but untranslated genes called long noncoding RNAs (lncRNA). First disregarded as cellular byproducts, functional lncRNAs affecting whole-body traits have been reported recently.

Given the intriguing potential of exercise (Review Paper) and the noncoding part of the human genome for gene-target discovery in age-related diseases, the goal of this thesis was: to test whether exercise alters expression of genes in heart failure, which could be harnessed to protect the failing heart from hypoxia (Paper I); to identify metabolic enhancer elements that alter activity upon diabetes-induction and harbor genetic variants associated with metabolism (Paper II); and finally, to dissect the intersection of exercise and the noncoding genome by searching for an exercise-altered lncRNA in skeletal muscle and characterizing such a lncRNA in muscle ageing (Paper III).

In Paper I, we discover PRODH expression to be reduced by heart-failure and rescued by exercise. PRODH is indispensable for maintaining mitochondrial bioenergetics and ATP levels in a hypoxic environment, while reconstitution of PRODH attenuates hypoxic impairments. In Paper II, we map skeletal muscle enhancer elements that alter activity upon diabetes induction and harbor single nucleotide polymorphisms associated with metabolic traits. Assessing physical interaction and transcriptional regulation between these enhancers and their target genes reveals several candidate genes whose expression correlate with whole-body metabolic traits. In Paper III, exercise-altered lncRNAs in skeletal muscle are identified. In worms, mice and humans the exercise-induced, pro-myogenic lncRNA CYTOR improves aged muscle function by reversing the age-associated loss of fast-twitch, type II muscle fibers. In conclusion, harnessing exercise and the noncoding genome for novel gene target discovery identified several promising gene targets implicated in the age-related diseases heart failure, diabetes and sarcopenia.

 

 

Bjørn Bredesen at University of Bergen

Trial lecture November 2nd at 10.15
Topic for trial lecture: “Spatial transcriptomics”
External link to the trial lecture is found here.

 

Dissertation: November 20th at 10.15-12.00

Title of thesis: “Modelling the structure, function and evolution of Polycomb/Trithorax Response Elements”

External link to dissertation here.

 

Sonja Lagström at University of Oslo

Trial lecture November 6th at 08.30
Topic for trial lecture:
“Tracking the COVID19 pandemics using next generation sequencing (NGS)  – sequence variants in the Nordic Countries.” External link to trial lecture here.
 
Dissertation: November 6th at 10.30
Title of thesis:“Characterisation of human papillomavirus genomic variation and chromosomal integration in cervical samples”
 
Summary:

Persistent infection with a high-risk human papillomavirus (HPV) type is necessary for cervical cancer development, causing nearly 5% of all cancers worldwide. Nevertheless, only a small fraction of HPV infections progress to cancer, indicating that additional molecular factors contribute to the development of cervical cancer.

The thesis aimed to characterise and explore mutations in the HPV genome and viral integrations into the human genome contributing to HPV-induced carcinogenesis. This can reveal new insight into cervical cancer development.

A unique next-generation sequencing protocol, TaME-seq, was developed for analysis of HPV genomic variation and integration. The results show that the overall intra-host HPV genomic variability is higher than previously assumed, with a high number of HPV genome variants found in all samples from early infections to cancer. A noticeable part of the mutations in HPV16, which is the most carcinogenic HPV type, was associated with the APOBEC3-enzyme that is suggested to be involved in viral clearance. The findings revealed integration sites that located both in previously reported and novel genomic sites. A large number of integrations was observed in or close to human cancer-related genes, which could be an indication of a more aggressive infection.

The TaME-seq method could potentially be a valuable method for assessing the risk of developing cervical cancer. An additional HPV screening test would enable more personalised follow-up, improving detection of lesions with higher risk of progression and reducing unnecessary follow-up and treatment of women with minimal risk of developing high-grade lesions or cancer.

 

Yaxin Xue at University of Bergen

Trial lecture October 26th at 13.00
Topic for trial lecture: “Computational challenges and approaches linked to use of long-read sequencing in metagenomics”.
 
Dissertation: November 5th at 10:15-13:00
Title of thesis: “Development and application of computational methods for NGS-based microbiome research”
 
 
Xiaokang Zhang at University of Bergen
Trial lecture October 19th at 13.15-14.00
Topic for trial lecture: “Machine learning approaches in personalized medicine”.
External link to the trial lecture is found here.
 
Dissertation: October 30th at 10:15-13:00
Title of thesis: Biomarker Discovery Using Statistical and Machine Learning Approaches on Gene Expression Data
 
Summary:
Cod is an important fish for Norway and is used as a model organism to learn about how environmental toxicants affect biological systems. In the dCod1.0 project, we have studied how fish react to toxicants at molecular level. Using sequencing technology, we have measured the expression of several thousand genes in liver samples from cods exposed in the laboratory or from contaminated environments. An interesting question is which genes are activated when the fish are exposed to environmental toxicants.
 
One technology for measuring gene expression is RNA sequencing. The data that is produced must go through a series of steps to obtain gene expression. There are many tools to automate this process. However, many of them are made for special applications and for data from model organisms such as humans or mice. Therefore, we have developed a workflow called RASflow that can be easily used without special programming skills and with different research interests.
 
When the gene expression profiles are clear, both traditional statistical hypothesis testing and machine learning methods can be applied to find out the individual genes or gene sets that show a reaction to the toxicants. The performance of the individual method is very dependent on the data. In addition, the methods are often unstable when the number of samples is low and the number of genes is high. Motivated by this, we developed a framework that makes it possible to combine different methods to identify relevant genes and that shows stable behavior across all the data sets we have analyzed.
 
External link to defense here.
 

Joseph Diab at University of Tromsø – The Arctic University

Time: October 9th 10:15-16:00

Thesis title: “The Metabolome and Lipidome of Ulcerative Colitis”

Summary of the thesis:

Inflammatory bowel disease (IBD) is a chronic, relapsing inflammatory disorder in the gastrointestinal tract affecting up to 0.5% of the population of the Western world. The two major forms of IBD, Ulcerative Colitis (UC) and Crohn’s Disease (CD), are characterized by a dysregulated immune response triggered by several genetic, microbial and environmental factors.

We preformed mass spectrometry-based metabolomic and lipidomic analysis on colon biopsies from UC patients to unravel the disease pathobiology and to identify markers for the disease outcome. We found that the alteration in lipid mediators correlates with the severity of UC. Moreover, we report potential prognostic and diagnostic markers for UC, such as very long chain ceramids and lipid mediators. Likewise, tryptophan metabolism is a key aspect of the impaired metabolism in active UC.

This work demonstrates the importance of metabolomics in IBD to identify key drivers of pathogenesis which prerequisite personalized treatment.

The trail lecture starts at 10.15, the defense starts at 12.15

Trial lecture:
Title:  “Proteomics in Inflammatory Bowel Disease – Integrated multi-omics to facilitate personalized medicine in IBD”

Link to the trial lecture and defense at UiT.

 

Zhi Zhao at University of Oslo
Time: October 9th 10:15-16:00
Thesis title: “Multivariate structured penalized and Bayesian regressions for pharmacogenomic screens”
 
Summary of the thesis:

Pharmacogenomic screens for personalized cancer therapy are the focused biomedical application in this thesis. Due to the complex relationships between targeted cancer drugs and high-dimensional genomic predictors, we have developed penalized likelihood methods and Bayesian hierarchical models to capture the complex structures in the pharmacogenomic data and to predict drug sensitivity.

The first part of the thesis proposed to address the correlations between drug sensitivity measures for multiple cancer drugs and the heterogeneity of multiple sources of genomic data in multivariate penalized likelihood methods with structured penalities. The proposed methods can improve the prediction performance of drug sensitivity. The second part of the thesis exploited Bayesian priors for the relationships between multiple drugs and relationships between drug sensitivity and the targeted pathways or genes of cancer drugs. Large pharmocogenomic screens may also include samples from multiple cancer tissue types. We employed random effects to address the sample heterogeneity in the proposed Bayesian model. The results have shown good structure recovery in the complex data and good prediction of responses by the new Bayesian models.

We congratulate NORBIS PhD member Zhi Zhao on his upcoming dissertation! Tomorrow (Friday October 9th) at 12.15 he will defend his thesis “Multivariate structured penalized and Bayesian regressions for pharmacogenomic screens” at the University of Oslo. The trial lecture starts at 10.15. Both trial lecture and thesis defense will be streamed online and all who like can join in. Thesis abstract and (external) links to trial lecture and defense are found here:

https://www.med.uio.no/imb/english/research/news-and-events/events/trial-lectures/2020/zhao-zhi.html
https://www.med.uio.no/imb/english/research/news-and-events/events/disputations/2020/zhao-zhi.html