240006 Bioinformatics 2

Details
Department of Bacic Animal and Veterinary Sciences   35 %
Department of Veterinary Pathobiology   15 %
Department of Natural Sciences   25 %
Department of Plant Biology   25 %
Earliest Possible YearMSc. 1 year to MSc. 2 year
DurationOne block
 
Credits7.5 (ECTS)
Course LevelMSc
 
ExaminationFinal Examination

written examination


All aids allowed

Description of Examination: Evaluation of a report based on the students own data or data provided from one of the teachers. A report is turned in by the end of the course.

Weight: Report 100%



13-point scale, internal examiner
 
Organisation of TeachingLectures, exercises
 
Block PlacementBlock 2
Week Structure: B
 
Teaching LanguageEnglish
 
Optional PrerequisitesBioinformatik1, Statistik, Matematik og Databehandling, Genetik ell. tilsvarende
 
Areas of Competence the Course Will Address
Scientific:
The student should aquire knowledge not only on how to apply methods available on the internet, but also on how to integrate them in biological work.

Technology and production:
Furthermore the student should also be able to on their own to obtain knowledge of new methods and understand how to apply them.

Ethics and values:
To relate and communicate research results in a critical way.
 
Course Objectives
The goal of the course is obtain basic bioinformatics knowledge at a level which include
machine learning methods and methods at a similar
level. The course will together with bioinformatics1
complete introduction of basic concepts in bioinformatics.

 
Course Contents
Phylogeny and sequence evolution:
Phylogenetic methods are based on multiple sequence alignments and can
be used to group molecular sequences and to reconstruct evolutionary
history. We introduce different models of molecular evolution and
present principles and methods for phylogenetic analysis. The
students will learn how to construct phylogenetic trees and to
critically judge them, using knowledge about the limitations of
phylogenetic analysis from a biological and from a statistical point
of view.


Neural networks (Machine learning part I):
We demystify neural networks, examine their basic principles, and
describe their application to prediction of protein secondary
structure as well as some of their many other applications. The aim
is that the student understands the basic principles and can determine
when a meaningful and reliable prediction has been provided.


Hidden Markov models (Machine learning part II):
Hidden Markov models (HMMs) are also widely used in bioinformatics.
We show their application to gene finding, promoter identification
(motif recognition), modelling of families and prediction of protein
secondary structure. The student should understand the principles and
assumptions of hidden Markov models to critically judge the
appropriateness of this class of models in a given situation and to
evaluate their use in algorithms and software.


Gene finding:
A fundamental task in genome sequencing projects is gene finding,
because knowledge of the genes is a prerequisite for downstream
analyses such as function, structure, and metabolism. We present
several strategies for gene finding in anonymous DNA sequences,
e.g. ab initio strategies, similarity driven strategies, and EST-based
strategies. Likewise, we present methods (e.g. neural networks, HMMs,
statistics-based) that implement the strategies, and discuss their
their strengths and weaknesses. The lectures will mainly be oriented
towards the practical use of the methods, as several of them are
poorly documented.


Non-coding RNA genes and RNA structure (RNA informatics):
Non-protein coding RNAs (ncRNAs) have turned out to be a highly abundant class
of genes. They play a central role in regulation of protein coding genes and
can be expressed at specific developmental stages or in specific tissues.
Furthermore, ordinary mRNA can contain structural elements in their
untranslated regions (UTRs) which may be involved in cis regulation of coding
genes. Various examples of ncRNAs are introduced along with methods for
predicting RNA structure and ncRNA genes.


Microarrays (2):
Microarrays play a steadily increasing role in simultaneously
analyzing the expression of tens of thousands of genes in a series of
biological and medical applications. We study topics from
Bioinformatics 1 in more depth and introduce new analysis methods. We
present the steps in image scan analysis, quality control and
normalization methods, and we discuss experimental designs such as
reference designs, loop designs and designs for time-course
experiments. Statistical methods include finding significant genes,
class prediction by discriminant analysis, class discovery by cluster
algorithms and analysis of pathways.


Metabolomics:
Metabolomics deals with the analysis of metabolites and can be used to
classify transgenic organisms and to investigate the effects of
external impacts, such as drug delivery, at the metabolite level. We
introduce techniques for metabolite analysis and statistical methods
for calculating differences of chromatographic profiles as well as
methods for analyzing the relationships between metabolite data and
phenotype/genotype. Metabolic networks will also be covered.


Details can be found at http://bioinf.kvl.dk/bioinformatics2
 
Teaching And Learning Methods
Lectures and exercises with supervision as well as group work with exercises. Case work for example in relation to a scientific paper can also take place.
 
Course Coordinator
Jan Gorodkin, gorodkin@genome.ku.dk, Department of Basic Animal and Veternary Sciences/Genetics & Bioinformatics, Phone: 35333578
 
Study Board
Study Committee NSN
 
Course Scope
lectures35
practicals40
project work80
preparation36
supervision5
theoretical exercises10

206