The rapid and reliable detection of pathogens is of utmost importance in healthcare settings to ensure the appropriate treatment thereby reducing morbidity and mortality for the patient. Current culturing, PCR based and NGS species detection methods are time consuming (Opota et al., 2015), limited in their detection (Buckley et al., 2015), or require specialist skills and are expensive (Basho and Eterovic., 2015). Oxford Technologies Nanopore devices could provide detailed genomic sequencing at a fraction of the cost and without the need for technical bioinformatic skills. This study evaluates the MinION device and analysis tools to suggest best practice. Classification and genotyping of 12 Klebsiella isolates were performed using EPI2ME automated workflows and manual de novo assembly.  Automated workflows using raw MinION reads provided clinically relevant information identified in ~6hrs. Manual de novo assembly and analysis used hybrid, and single source data took >24hrs. The inclusion of MinION long reads overcome problems assembling short reads. Hybrid genomes provided the most contiguous and highly detailed contigs. MinION only read assemblies contained more errors but still identified similar genotypic findings. Automated workflows are rapid and require minimal bioinformatic know-how. There should be a dialogue between clinicians and bioinformaticians to develop bespoke analysis tools.  Although challenges remain around compatible kits and vulnerable flowcells long read sequencing can be an effective tool for species detection and pathogen typing. Furthermore, hybrid assemblies have the potential to advance our genome detailing and discovery.

The widely collected and analyzed genetic data help in understanding the underlying mechanisms of heterogeneous diseases. Cellular components interact in a network fashion where genes are nodes and edges are the interactions. The failure in individual genes lead to dys-regulation of sub-groups of genes which causes a disease phenotype, and this dys-functional region is called a disease module. Disease module identification in complex diseases such as asthma and cancer is a huge challenge. Despite the development of numerous sophisticated methods there is a still no gold standard. In this study we apply different parameter settings to test the performance of a widely used method for disease module detection in multi-omics data called Weighted Gene Co-expression Network Analysis (WGCNA). A systematic approach is used to identify disease modules in asthma and arthritis diseases. The accuracy of obtained modules is validated by a pathway scoring algorithm (PASCAL) and GWAS SNP enrichment. Our results differ between the tested data sets and therefore we cannot conclude with recommendations for an optimal setting that could perform best for multiple data sets using this method.

Oats (Avena sativa) have been one of the most important cereals in Swedish crop history, completely dominating domestic cereal production in the 19th and early 20th centuries. During this time, oats were mainly used as horse feed and since then, production has decreased with decreased use of horses. Oats have recently been rediscovered as a source of plant-based protein with several health benefits, and due to the low environmental requirements and effectiveness in cereal crop rotation, oats are yet again increasing in popularity. For future development of oat cultivars, insight into oat genetics and genomic history of oats is required. In this study, 84 archeological oat grains (1440-1766 AD) originating from desiccated and waterlogged sources in southern Sweden were analyzed to understand their genetic history. Next generation sequencing (NGS) of extracted DNA from individual grains revealed that the samples had varying oat alignment success, ranging from 0.01 - 34.38% (average: 9.00%). Through metagenomic analysis, it was concluded that, even though a large part of the acquired reads were bacterial, the endogenous oat content was enough to enable deeper bioinformatic analysis. Through Kompetitive Allele Specific PCR (KASP) genotyping, the genetic and geographic structuring was investigated by comparison with 343 previously genotyped 19th century oat landrace grains. Results showed that ancient landraces share genetic similarity with historical landraces and modern improved cultivars, implying that the with the introduction of improved cultivars, they were incorporated into ancient landraces, possibly to increase local adaptation. The analyses revealed a large within-accession variation in all results. An attempt was made to understand the source of this variation, but no pre-extraction parameters that increase sample success were distinguishable. With this study, it was shown that genotyping, sequencing and bioinformatic analysis of ancient oat samples is possible, and can be used to understand oat genetic history.

Arbuscular mycorrhizal (AM) fungi are obligate biotrophs that formsymbiosis with plants by colonizing their roots. The fungus supplies thehost plant with various nutrients, including phosphorus. Phosphorus iscrucial for the development of plants and is hard to acquire in soilsince it can be scarce and has a slow motility. The fungus utilizes itslong hyphal threads to contact more soil to obtain phosphorus andtransport it back to the plant. It does so with its use of differentphosphorus transporters (PTs) located in its membranes. Here we havedone a phylogenetic analysis of different PTs from a variety of fungifrom different phyla together with plants and new sequence data from AMfungi. In total, 955 genomes were screened, 26 of which belong to AMfungi. This work resulted in a database of 1351 PT sequences, 907 fromfungi (243 from AM) and 444 from plants, and two phylogenetic trees tovisualize the data. One phylogeny was made of the branch of the PT Pho87which was selected for building a Hidden Markov model, which canfacilitate future searches of PTs.

The scope of this project was to implement a quorum sensing model capable of synchronised oscillations from the article ”A synchronized quorum of genetic clocks” [1] in the software framework URDME [2]. The model consists of a system of partial differential equations describing time delayed and coupled biochemical reactions. In URDME, the time delay system was formed using a cascade of reactions in which the rate of each reaction was set so that the expected total time for all reactions in the cascade corresponds to a certain delay time. One reason for this cascade delay model is that it might better capture the inherently stochastic nature of the delay mechanism in the quorum sensing network, as opposed to a model using explicit delays.Another reason is simplicity of implementation, as delays are not explicitly supported in URDME. After initial tests suggested that the cascade delay model gave satisfying results, it was incorporated into the quorum sensing model from the article, which was implemented by rewriting the differential equations as a system of biochemical reactions. Simulations in one and two dimensions were then done, with both stochastic and deterministic solution methods. The one dimensional and two dimensional simulations yielded distinct synchronised oscillations with a cascade delay containing five sub-reactions. Several results from the simulations of the original article could be reproduced. From the results, it was concluded that the proposed cascade delay model was successful in modelling the delayed reactions in the quorum sensing network. In future studies, it is suggested that the individual cells, in which most of the reactions in the quorum sensing network happen, are modelled with greater resolution.

Cattle are essential for the global food industry through the meat and milk production. It is from an economical point of view in our best interest to make cattle as efficient as possible, whether it is milk or beef production, without negatively influencing their health and welfare. That has led to a steady increase in the interest of genetic analysis of cattle. The sequencing and identification of genomic variation has led to the association of genotypes with phenotypes of interest and the discovery of the underlaying genetic risk factors for many diseases and traits. Diseases or monogenetic traits caused by a single nucleotide polymorphism (SNP), small deletions and insertions or other small mutations are often easy to identify if the correct region is found. The diseases caused by structural variants (SVs), variants larger than 50 base pairs (bp) are still challenging. It is more challenging because they are harder to identify, especially using shortread sequencing technologies. It is therefore still a rather unexplored area for cattle and other domestic species.This thesis looks at SVs found in the Swedish Red and Brown (SRB) cattle to discover breed specific SVs. This was done by creating a pipeline with VCF files as input. The identified SVs were filtered and overlapped with externally identified SVs. The pipeline was tested with two SRB datasets. The structural variant caller, DELLY, performed poorly with low read depth data when comparing single replicate data and combined replicates data. Multiple SVs were identified in all individuals and did overlap with both functional and gene annotation. There was also overlap found with datasets in the European variant archive (EVA). This indicates that the identified SVs are shared among multiple breeds of cattle and that DELLY can be used to develop future pipelines to include long read sequencing technologies and/or data with higher read depth.

The analysis of kinship from ancient DNA (aDNA) data has the potential to provide insight into social structures of prehistoric societies. Kinship analysis is gaining popularity as optimised wet-lab methods allow for studies with sample sizes on the level of whole cemeteries. However, the specifics of ancient DNA require different methods than what would be used for modern DNA. A common way is to use the sites that are identical-bydescent (IBD), however, detecting these is often a challenging task since it is not easy to determine whether a shared locus between two individuals is inherited from the ancestor or if another factor caused the similarity. Most methods used in the field are able to identify up to 2nd or 3rd degree relatives from aDNA data but do not distinguish between different types of relationship for the same degree, for instance not being able to differentiate between parentoffspring and full sibling-sibling relationship in first degree. The aDNA kinship methods often use either of window-based or single-site approaches, however, these two approaches have not been compared formally before in terms of effectivity and efficiency. In this work, READv2 is presented as a re-implementation of a popular kinship analysis method for aDNA studies with additional features such as accepting .bed files as input, which take up less space than the previous input type, plain-text .tped files. It is shown that the new version works more efficiently in terms of runtime. However, the memory requirements seem to be increased with the new implementation. Furthermore, a window-based approach is compared with the single-site approach of READv2, as well as varying window sizes, with benchmarked simulation data which contains approximately 700 individuals with known 1st degree, 2nd degree and 3rd degree relationships. According to the comparison, the sensitivity of the method does not vary between the approaches and different window sizes for high coverages. However, the single-site approach has been shown to be the superior one by a small margin for lower coverages. In addition to these, using the variance of non-shared alleles in windows along the genome has been used to implement a method to differentiate different first-degree relationships, parent-offspring and siblings. The method is tested with an independent dataset from the 1000 Genomes Project which shows that the proposed method is able to work with different datasets with varying sets of SNPs. Nevertheless, the first-degree classification method requires further analyses to determine the stress-point where the True Positive rates for both categories start to drop. Additionally, some necessary changes and decisions are required for READv2 to be a user-friendly method that can be used by other researchers. The preliminary release of READv2, including example data as well as instructions to install the necessary packages and to run the algorithm can be found in https://github.com/GuntherLab/READv2/releases/tag/READ.

Predicting protein structures has long been an area of active research in the field ofbioinformatics. Great strides have recently been made in this area by googles DeepMindteam. They developed an AI called AlphaFold which is able to make the most accuratepredictions of protein structures as of date. With the advent of AlphaFold some considerthe problem solved. There is however an area in protein prediction that has lagged be-hind, that of multi conformational prediction. There are proteins that can take on oneout of several active forms in the body. Making predictions for these are harder than forsingle conformational proteins due to an increase in complexity and a lack of data. Apromising solution to this problem is to introduce noise to the input data AlphaFold usesto create a wider range of predictions. In this thesis multi conformational prediction withdifferent methods to introduce noise is evaluated. Dropout, disclosing templates, untar-geted Multiple sequence alignment(MSA) subsampling and targeted MSA subsamplingwere used. It was concluded introducing noise did indeed improve the prediction of mul-tiple conformations. Among them, MSA subsampling seemed to be the most effective,especially untargeted MSA subsampling. Dropout also seemed to slightly improve theresults while excluding template information did little to nothing. AlphaFold was unableto predict both structures for 6 out of 16 structures, even with introduced noise. No clearreason for why this could be determined, but the leading hypothesis is that AlhpaFoldwas unable to extract sufficient information about both conformations from the MSAdata for these proteins.

This project aimed to find a method for producing potential probe targets for identification of ESBL (Extended Spectrum Beta Lactamase) genes in bacteria. ESBLs are a type of enzymes responsible for antibiotic resistance in many bacteria. The result we developed was a semi-automated pipeline that utilises several Perl scripts to download gene sequences, identify sequence subgroups based on sequence similarity, find common target sequences among them and screen the target sequences against a background database. These target sequences should work with padlock probes and therefore had specific requirements regarding length and highest number of allowed mismatches. This report includes descriptions of the scripts and ideas for future improvements, as well as an ethical analysis about aspects relevant to research on antibiotic resistance.

Metagenomic sequencing is an increasingly popular way of determining microbial diversity from environmental and clinical samples. By specifically targeting the 16S rRNA gene found in all bacteria, classifications of pathogens can be determined based on the variable and conserved regions found in the gene. Metagenomic sequencing can therefore highlight the vast difference in microbiological diversity between culture-dependent and culture-independent methods. Today, this has expanded into various next-generation sequencing platforms which can provide massively parallel sequencing of the target fragment. One of these platforms is Ion-torrent, which can be utilized for targeting the 16S rRNA gene and with the help of bioinformatics pipelines be able to classify pathogens using the bacteria’s own variable and conserved regions. The overall aim of the present work is to evaluate the clinical use of Ion-torrent 16S ribosomal RNA sequencing for determining pathogenic species from clinical samples, but also to set up a pipeline for clinical practice. Optimal DNA-extraction and quantification methods were determined towards each evaluated sample-type and DNA-eluates were sent for 16S rRNA Sanger and Next-generation sequencing. The result indicated that the next-generation sequencing shows a concordance in results towards the culturing-based method, but also the importance of experimental design and effective quality trimming of the NGS data. The conclusion of the project is that the Ion-torrent pipeline provided by the Public Health Agency of Sweden shows great promise in determining pathogens from clinical samples. However, there is still a lot of validation and standardisations needed for the successful implementation into a clinical setting.