Wait a second...
Nepřihlášený uživatel
iduzel: 52782
idvazba: 60825
šablona: api_html
čas: 7.6.2020 11:42:44
verze: 4671
uzivatel:
remoteAPIs: https://cis-web.vscht.cz/zaverecne-prace/pracoviste/
branch: trunk
Obnovit | RAW

Department of Informatics and Chemistry

List of available PhD theses

Analysis of expression of endogenous retroviruses in human and mouse thymus

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programme: Bioinformatika
Theses supervisor: Mgr. Jan Pačes, Ph.D.

Annotation

Endogenous retroviruses comprise millions of discrete genetic loci integrated into the genomes of all vertebrates. These genetic loci represent past retroviral infections and their ability to integrate into the chromosomal DNA of germ-line cells has endowed retroviruses with the potential to be fixed in the genomes during evolution. Despite the fact that endogenous retroviruses lost this ability to infect other cells, presence of strong promotors in their sequences allows them to express some of their genes in host cells.

In this work, we study gene expression of human and mouse ERVs in the thymus in context of autoimmune diseases. We analyze RNAseq data and compare expression patterns of ERVs in thymus to those in other tissues. Single cell RNAseq data from mTECs is analyzed to obtain detailed expression patterns. Our aim is to assess to what extent are HERVs included in the presentation of self antigens during negative selection of T cells and discuss what are the implications in autoimmune disorders. Results from this diploma work will be integrated into the internationally acclaimed database of endogenous retroviruses HERVd.

Application of Advanced Molecular Methods to Reveal Bacterial Mechanisms of Virulence

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology

Annotation

Classical Bordetella species cause respiratory infections of mammals, such as the whooping cough disease of humans caused by Bordetella pertussis and B. parapertussis or kennel cough in dogs caused by B. bronchiseptica. These pathogenic bacteria employ a type III secretion system (T3SS) to inject cytotoxic BteA effector protein into cells of the mammalian hosts. It remains unknown how BteA effector protein functions and contributes to the diverging biology of classical Bordetella species. The aim of this thesis is to use genome-wide CRISPR/ Cas9-based screening to identify genes confirming susceptibility to BteA cytotoxicity. The PhD. candidate will perform lentiviral transduction of CRISPR/Cas9 gRNA library followed by Bordetella infection and next-generation sequencing of total DNA isolated from surviving cells. After data analysis hits will further be validated and mapped in signaling and biosynthetic pathways.

Application of Advanced Molecular Methods to Reveal Bacterial Mechanisms of Virulence

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology

Annotation

Classical Bordetella species cause respiratory infections of mammals, such as the whooping cough disease of humans caused by Bordetella pertussis and B. parapertussis or kennel cough in dogs caused by B. bronchiseptica. These pathogenic bacteria employ a type III secretion system (T3SS) to inject cytotoxic BteA effector protein into cells of the mammalian hosts. It remains unknown how BteA effector protein functions and contributes to the diverging biology of classical Bordetella species. The aim of this thesis is to use genome-wide CRISPR/ Cas9-based screening to identify genes confirming susceptibility to BteA cytotoxicity. The PhD. candidate will perform lentiviral transduction of CRISPR/Cas9 gRNA library followed by Bordetella infection and next-generation sequencing of total DNA isolated from surviving cells. After data analysis hits will further be validated and mapped in signaling and biosynthetic pathways.

Bioinformatic studies of G-quadruplex biochemical specificity

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika
Theses supervisor: Edward A. Curtis, Ph.D.

Annotation

G-quadruplexes are noncanonical nucleic acid structures thought to play widespread biological roles. The algorithms currently used to identify G-quadruplexes in genomes use models in which structurally distinct classes of G-quadruplexes, such as those with different strand polarities, are grouped together. However, emerging evidence suggests that these models are too general because in some cases they cannot distinguish G-quadruplexes with biochemically distinct functions. To address this limitation, here we will perform a series of functional screens using G-quadruplex libraries in which parameters such as tetrad number, loop length, and loop sequence are systematically varied. The data from these screens will be used to generate sequence models for G-quadruplexes specific for particular functions. The genomic distribution of these classes of G-quadruplexes will then be analyzed with respect to annotated genomic features such as transcription start sites. We hypothesize that this analysis will reveal genomic associations that are currently undetectable due to noise from the more general G-quadruplex models currently used.

Deep learning for de novo molecular design

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programme: Bioinformatika
Theses supervisor: doc. Mgr. Daniel Svozil, Ph.D.

Annotation

De novo molecular design aims to generate novel chemical compounds with desirable chemical and pharmacological properties from scratch using computer-based methods. Recently, deep generative neural networks have become a very active research frontier in de novo molecular discovery. Though these methods enable the unprecedented coverage of chemical space, many aspects of their application still remain to be studied in a thourough detail. The dissertation will focus on various aspects of deep molecular generators including molecular representation, the creation of compounds with desired properties or the benefits of new deep architectures.

Development of machine learning algorithms for the prediction of terpene synthase activity.

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika
Theses supervisor: Ing. Tomáš Pluskal, Ph.D.

Annotation

Bioactive plant metabolites are an essential source of chemical scaffolds for the development of new medicines. Our group is interested in establishing new cutting-edge methods to harness the biosynthetic potential of plant specialized metabolism for the benefit of human health. One of the interesting problems is prediction of enzyme activity from its amino acid sequence using computational approaches. In this project, we will develop machine learning platforms (e.g., deep neural networks) for predicting the chemical structures produced by a specific class of plant metabolic enzymes, terpene synthases. These enzymes are collectively responsible for generating the largest and the most diverse family of plant specialized metabolites with numerous medical and industrial applications. As side projects, the student will contribute to the development of computational workflows for mass spectrometry, metabolomics, molecular networking, and de novo transcriptome sequencing. References: 1. Christianson, D. W. Structural and Chemical Biology of Terpenoid Cyclases. Chem. Rev. 117, 11570–11648 (2017) 2. Vattekkatte, A., Garms, S., Brandt, W. & Boland, W. Enhanced structural diversity in terpenoid biosynthesis: enzymes, substrates and cofactors. Org. Biomol. Chem. 16, 348–362 (2018)

High-throughput sequencing data processing for taxonomic and (meta)genomic analyses in microbial ecology

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programme: Bioinformatika
Theses supervisor: doc. Ing. Ondřej Uhlík, Ph.D.

Annotation

With the advent of molecular-biological methods and new technologies such as high-throughput DNA sequencing, microbial ecologist have been enabled to study in detail the composition of microbial communities without the need of complicated isolation of individual species. These analyses are referred to as metagenomics and rely on the analyses of all genomes contained within an environmental sample. Application of these methods allow for analyzing of habitats which have not been chracterized previously. In the framework of this thesis, the student will analyze high-throughput sequencing data for the sake of taxonomic and (meta)genomic analyses, including the application of currently used as well as development of novel bioinformatics tools and pipelines with the aim to characterize microbial communities in soils and extreme habitats.

High-throughput transcriptomics platform

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika
Theses supervisor: RNDr. Petr Bartůněk, CSc.

Annotation

The aim of the project is to develop a robust ultra-high-throughput sequencing platform and a bioinformatics pipeline for the analysis of gene expression in chemical biology and drug discovery. Platforms that are currently used in academic research and pharmaceutical development are limited by their high cost and low throughput. Therefore, we will establish a cost-effective high-throughput transcriptome profiling strategy with simplified sample preparation and multiplexing in 384/1536-well format. The transcriptional profiles will be correlated with the Mechanism of Action (MoA) of the analyzed compounds, which will facilitate mechanistic studies of novel compounds and their biological activities. The scheme can be combined with the CRISPR gene-editing methods to enable the discovery of gene functions in cells and organisms.

Integration of phenotyping and functional genomic data

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika
Theses supervisor: Vendula Novosadoivá

Annotation

The position of bioinformatician is becoming necessary for every scientific group. Generating large datasets of omic data makes it necessary to develop new computational algorithms using tools such as machine learning and artificial intelligence, which will also allow the processing of diverse unstructured data. Our group is part of the research infrastructure Czech Centre for Phenogenomics, involved in the systematic annotation of the mouse genome within the International Mouse Phenotyping Consortium (IMPC). We produce mouse lines with one gene deactivated. These lines are further characterized by a standard phenotyping pipeline. The data set from each animal tested has over 700 parameters from different fields. These parameters contain numeric, categorical and image data. We are also collecting metabolomic data for selected lines. The Ph.D. project aims to integrate every data generated both in our center and within the whole IMPC. Linking individual parameters and finding correlations and causality between them and their possible semantic analysis will help to better understand the phenotype. At the same time, knowledge of a given gene function will enable mathematical modeling of the phenotype of genes involved in similar or overlapping regulatory networks.

Microbial community patterns under soil land use changes.

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika

Annotation

Soil is considered the most microbial diverse environment on Earth. Understanding the traits that control the biological entities in such rich environment is of interest to all fields of life sciences, from ecology and environmental microbiology to agriculture, biotechnology and health. Effects of human activities such as land use changes had led to temperature increase, nitrogen contamination, with a direct impact on microbial biomass and their functional profiles. The current methods for studying microbial communities produce data that allows us to identify latent variables which can control the changes in microbial communities, considering the power of sequencing the whole community and disentangling sample relationship from the environmental data. The limitation nowadays resides in gathering, processing and analyzing the sample data in a standardized way, on top of the specialist interpretation of the results. The proposed solution to such a problem would be to build a community fed database that could be used together with a pipeline for identifying those latent variables, which would be the species response to environmental changes.

Modelling elastic properties of nucleic acids structures

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programme: Bioinformatika
Theses supervisor: doc.Ing. Filip Lankaš, Ph.D.

Annotation

Elastic properties of molecules determine free energz changes upon their mechanical deformation. Elasticity (or deformability) of nucleic acids plays a key role in their recognition by proteins, affects spatial organization of the genome or, in the case of RNA, is of utmost importance in the functioning of the ribosome and other macromolecular complexes. The aim of the work is to develop models describing the elasticity of DNA and RNA structures and to parameterize the models from extensive molecular dynamics simulations. Systems important in biology and for nanotechnological applications will be chosen, such as DNA duplexes of various sequences, DNA containing a radiational damage, or various RNA structural motifs. The work will contribute to our understanding of DNA and RNA mechanical properties and their role in the context of biology, biophysics and nanotechnology.

The role of PRDM9 histone methyltransferase in genetic recombination and reproductive isolation between species

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika

Annotation

The Prdm9 gene determines localization of meiotic recombination hotspots in genome of mice, humans and other mammalian species. In our laboratory we discovered another function of Prdm9, as a major hybrid sterility gene responsible for infertility of hybrids between related mouse species. To verify Prdm9 as the first speciation gene in vertebrates various combinations of Prdm9 alleles will be tested by genome-wide mapping of PRDM9 hotspots and their correlation to fertility phenotypes. The optical maps at the large scale and the genomic hotspot maps of recombination will be generated by Bionano optical mapping technology and by Illumina platform based ChIP-seq technology, respectively. The aim of this PhD thesis will be, using bioinformatics tools, to analyze the acquired datasets with the perspective to get correlations between maps from mice with different phenotypes and to determine key components of genomic structure and recombination landscape crucial for hybrid sterility.

Transcriptome analysis of acute and degenerative disorders in the central nervous system

Department: Department of Informatics and Chemistry, Faculty of Chemical Technology
Study programmes: Bioinformatics, Bioinformatika
Theses supervisor: doc. Mgr. Daniel Svozil, Ph.D.

Annotation

In the last decades, our knowledge of the molecular basis of life has increased tremendously. Technical breakthroughs from physics, chemistry and biology facilitated the development of new technologies producing massive amounts of data. Particularly, RNA sequencing (RNA-Seq) has dramatically influenced the landscape of many life science disciplines. Although large RNA-Seq datasets can be produced in a rapid and cost-effective way, their analysis, visualization and interpretation represent a major bottleneck in the future development. The main focus of this thesis is to develop and implement computational pipelines for analysis of various RNA-Seq datasets generated in acute (stroke, spinal cord injury) and degenerative disorders (amyotrophic lateral sclerosis and Alexander disease) in the central nervous system. The spectrum of datasets encompasses standard bulk RNA-Seq data, so as datasets of small RNA species and single-cell transcriptomes. As the RNA-Seq is a fast developing field, main emphasis will be put on the development, application and improvement of new algorithms and bioinformatics tools covering various steps of data analysis, including raw data processing, quality control, normalization, data visualization and integrative analysis.


UCT Prague
5 Technická
166 28 Prague 6 – Dejvice
IČO: 60461337
VAT: CZ60461373

Data mail: sp4j9ch

Copyright UCT Prague 2014
Informations provided by Department of International Relations, Department of R&D, technical service by Computer Centre
switch to desktop version