ZAIK - Gruppe Faigle/Schrader: Projekte/Bioinformatik


ZAIK - Group Faigle/Schrader: Research Transfer/Projects/Bioinformatics

Bioinformatics
Advances in experimental techniques have revolutionized the life sciences. Instead of merely describing phenomenological observations we are now in a position to begin to systematically analyze molecular mechanisms of many biological processes. This is partly due to the availability of the entire human genome sequence. The amount of DNA and Protein Sequences freely accessible in public databases is exponentially growing; today sequencing is not going on in small labs anymore, but in dedicated ,,Sequencing Factories``.
Bioinformatics is a new, independent, interdisciplinary research area, in which methods from mathematics, computer sciences and statistics play a crucial rule. One goal is to drive forward experimental techniques to lower costs - for example for the human genome project. This is done in close dialog with biologists, including biology, biochemistry, chemistry and physics knowledge. On the other hand, answering the relevant biological questions must be made possible by new approaches to data analysis and modelling.

Projects

Primer Design for Multiplexed Genotyping
The Polymerase Chain Reaction (PCR) is the workhorse of biotechnology. Multiplexing this reaction, and thus amplifying several DNA amplicons at different genomic loci simultaneously, can lead to significant time savings and cost decreases, and is thus of considerable interest for lab applications.
We have developed a computer program to assisst in the design of primers for the multiplexing of Polymerase Chain Reactions.
Lars Kaderali (kaderali@zpr.uni-koeln.de), Astrid Gösling, P Scott White, Rainer Schrader

A Fractional Programming approach to efficient DNA Melting Temperature Calculation
In many experimental techniques, the melting temperature of two given DNA strands is important. The selection of primers for the polymerase chain reachtion (PCR) or the design of oligonucleotide probes for DNA chips are examples, where efficient methods for the computation of DNA melting temperatures are required. We present a new computational method, based on Dinkelbach's fractional programming algorithm, that will simultaneously compute the most stable duplex and the corresponding melting temperature for two arbitrary (not necessarily complementary), given DNA strands.
Lars Kaderali (kaderali@zpr.uni-koeln.de), Alexander Schönhuth, Rainer Schrader in cooperation with Markus Leber (Institute of Biochemistry, University of Cologne).

Analysis of gene expression data using Hidden Markov Models DNA chip experiments have become routine in genetic network analysis. Performing microarray experiments consecutively in time produces time courses of gene expression levels. For modeling these time courses Hidden Markov Models have proven to be favourable. They allow

to integrate prior knowledge,
to visualize and analyze interactively and
they are robust with respect to noisy and missing data.
The current state of the art can be inspected here.
Alexander Schönhuth (aschoen@zpr.uni-koeln.de) together with Alexander Schliep and Christine Steinhoff of the Max Planck Institute for Molecular Genetics, Berlin.

Inference of gene regulatory networks based on gene expression data
A challenging problem in bioinformatics is the question how to find gene regulations of an organism. Such regulations can be described with the help of systems of differential equations. Therefore we use gene expression time-series data. The specification of the model as well as the methods to calculate the parameters of the differential equations are developed in our institute.
Jutta Gebert (gebert@zpr.uni-koeln.de) und Nicole Radde (radde@zpr.uni-koeln.de)

Pattern Recognition in Genetic Epidemiology
Methods of statistical pattern recognition carry the potential to analyse complex interactions in human genetic diseases. A focus of the work at the ZAIK is the development of new mathematical toos for this purpose.
Lars Kaderali (kaderali@zpr.uni-koeln.de)

Using Observable Operator Models to analyze biosequences
Statistical modeling of sequences offers a wide range of sequence analysis applications, some of which are model based clustering, pattern recognition and alignments. Theoretical research on the new class of Observable Operator Models (which can be understood as an extension of the well known Hidden Markov Models) as well as practical applications in gene expression analysis and protein classification is the subject of a Phd thesis at the ZAIK.
Alexander Schönhuth (schoenhuth@zpr.uni-koeln.de)

Finished Projects

Primer Design for Genotyping Applications
Single-base primer extension polymerase reactions can be combinet with DNA arrays or flow cytometry based techniques to permit strongly parallel analysis of multiple markers (multiplexing). Such an undertaking requires careful design of multiplexable SBE primers.
Lars Kaderali (kaderali@zpr.uni-koeln.de), Alina Deshpande, John P. Nolan, P. Scott White

ProClust: Protein Clustering - Searching for Homologue Proteins using transitivity
Analysis of protein sequences for protein structure prediction
Eva Bolten, Peter Piepenbacher, Alexander Schliep (schliep@molgen.mpg.de), Alexander Schönhuth (aschoen@zpr.uni-koeln.de, Sebastian Schneckener, Dietmar Schomburg, Rainer Schrader

Selecting probes for DNA Arrays
Massive parallel hybridization experiments as carried out on DNA Chips require careful selection of chip probes, in order to avoid cross-hybridizations. An Algorithm to do so has been developed in a masters thesis project.
Lars Kaderali (kaderali@zpr.uni-koeln.de), Alexander Schliep

Computation of the melting point of two DNA sequences
The melting point of two DNA-Sequences can be determed online. The provided form uses the programm ThermAlign.
Lars Kaderali (kaderali@zpr.uni-koeln.de)

Interest Group
The Bioinformatics Interest Group BIG is a loose group of employees of scientific and commercial institutions in the bioinformatics field. A key concern is to build scientific networks in the Cologne/Bonn region. To do so, meetings are held at regular intervals. Additionaly, two mailing lists have been created.
More information can be found at http://www.zaik.uni-koeln.de/~big

Partners

CUBIC Cologne University Bioinformatics Center
Science Factory
Arbeitsgruppe Genetik / Prof. Tautz
Institut für Biochemie / Prof. Schomburg
Los Alamos National Lab (LANL)

Contact

Contact via e-mail to bioinformatik@zpr.uni-koeln.de

Homepages Navigate Site Buttons

Bioinformatics	Advances in experimental techniques have revolutionized the life sciences. Instead of merely describing phenomenological observations we are now in a position to begin to systematically analyze molecular mechanisms of many biological processes. This is partly due to the availability of the entire human genome sequence. The amount of DNA and Protein Sequences freely accessible in public databases is exponentially growing; today sequencing is not going on in small labs anymore, but in dedicated ,,Sequencing Factories``. Bioinformatics is a new, independent, interdisciplinary research area, in which methods from mathematics, computer sciences and statistics play a crucial rule. One goal is to drive forward experimental techniques to lower costs - for example for the human genome project. This is done in close dialog with biologists, including biology, biochemistry, chemistry and physics knowledge. On the other hand, answering the relevant biological questions must be made possible by new approaches to data analysis and modelling.
Projects	Primer Design for Multiplexed Genotyping The Polymerase Chain Reaction (PCR) is the workhorse of biotechnology. Multiplexing this reaction, and thus amplifying several DNA amplicons at different genomic loci simultaneously, can lead to significant time savings and cost decreases, and is thus of considerable interest for lab applications. We have developed a computer program to assisst in the design of primers for the multiplexing of Polymerase Chain Reactions. Lars Kaderali (kaderali@zpr.uni-koeln.de), Astrid Gösling, P Scott White, Rainer Schrader A Fractional Programming approach to efficient DNA Melting Temperature Calculation In many experimental techniques, the melting temperature of two given DNA strands is important. The selection of primers for the polymerase chain reachtion (PCR) or the design of oligonucleotide probes for DNA chips are examples, where efficient methods for the computation of DNA melting temperatures are required. We present a new computational method, based on Dinkelbach's fractional programming algorithm, that will simultaneously compute the most stable duplex and the corresponding melting temperature for two arbitrary (not necessarily complementary), given DNA strands. Lars Kaderali (kaderali@zpr.uni-koeln.de), Alexander Schönhuth, Rainer Schrader in cooperation with Markus Leber (Institute of Biochemistry, University of Cologne). Analysis of gene expression data using Hidden Markov Models DNA chip experiments have become routine in genetic network analysis. Performing microarray experiments consecutively in time produces time courses of gene expression levels. For modeling these time courses Hidden Markov Models have proven to be favourable. They allow to integrate prior knowledge, to visualize and analyze interactively and they are robust with respect to noisy and missing data. The current state of the art can be inspected here. Alexander Schönhuth (aschoen@zpr.uni-koeln.de) together with Alexander Schliep and Christine Steinhoff of the Max Planck Institute for Molecular Genetics, Berlin. Inference of gene regulatory networks based on gene expression data A challenging problem in bioinformatics is the question how to find gene regulations of an organism. Such regulations can be described with the help of systems of differential equations. Therefore we use gene expression time-series data. The specification of the model as well as the methods to calculate the parameters of the differential equations are developed in our institute. Jutta Gebert (gebert@zpr.uni-koeln.de) und Nicole Radde (radde@zpr.uni-koeln.de) Pattern Recognition in Genetic Epidemiology Methods of statistical pattern recognition carry the potential to analyse complex interactions in human genetic diseases. A focus of the work at the ZAIK is the development of new mathematical toos for this purpose. Lars Kaderali (kaderali@zpr.uni-koeln.de) Using Observable Operator Models to analyze biosequences Statistical modeling of sequences offers a wide range of sequence analysis applications, some of which are model based clustering, pattern recognition and alignments. Theoretical research on the new class of Observable Operator Models (which can be understood as an extension of the well known Hidden Markov Models) as well as practical applications in gene expression analysis and protein classification is the subject of a Phd thesis at the ZAIK. Alexander Schönhuth (schoenhuth@zpr.uni-koeln.de)
Finished Projects	Primer Design for Genotyping Applications Single-base primer extension polymerase reactions can be combinet with DNA arrays or flow cytometry based techniques to permit strongly parallel analysis of multiple markers (multiplexing). Such an undertaking requires careful design of multiplexable SBE primers. Lars Kaderali (kaderali@zpr.uni-koeln.de), Alina Deshpande, John P. Nolan, P. Scott White ProClust: Protein Clustering - Searching for Homologue Proteins using transitivity Analysis of protein sequences for protein structure prediction Eva Bolten, Peter Piepenbacher, Alexander Schliep (schliep@molgen.mpg.de), Alexander Schönhuth (aschoen@zpr.uni-koeln.de, Sebastian Schneckener, Dietmar Schomburg, Rainer Schrader Selecting probes for DNA Arrays Massive parallel hybridization experiments as carried out on DNA Chips require careful selection of chip probes, in order to avoid cross-hybridizations. An Algorithm to do so has been developed in a masters thesis project. Lars Kaderali (kaderali@zpr.uni-koeln.de), Alexander Schliep Computation of the melting point of two DNA sequences The melting point of two DNA-Sequences can be determed online. The provided form uses the programm ThermAlign. Lars Kaderali (kaderali@zpr.uni-koeln.de)
Interest Group	The Bioinformatics Interest Group BIG is a loose group of employees of scientific and commercial institutions in the bioinformatics field. A key concern is to build scientific networks in the Cologne/Bonn region. To do so, meetings are held at regular intervals. Additionaly, two mailing lists have been created. More information can be found at http://www.zaik.uni-koeln.de/~big
Partners	CUBIC Cologne University Bioinformatics Center Science Factory Arbeitsgruppe Genetik / Prof. Tautz Institut für Biochemie / Prof. Schomburg Los Alamos National Lab (LANL)
Contact	Contact via e-mail to bioinformatik@zpr.uni-koeln.de