AG Lechner - Bioinformatics Junior Group
Department of Pharmaceutical Chemistry, Philipps-University Marburg
We operate at the interface of experimental and computational science.
As a result, algorithmic bioinformatics is as important as to apply bioinformatic techniques and tools in order to solve current questions in biology, biochemistry and pharmaceutical chemistry.
Our research mainly covers genomics, transcriptomics and molecular interactions.
A strong network of experimental and computational laboratories allows us to combine analyses
, in vitro
and in vivo
on a multitude of multidisciplinary subjects.
Our algorithmic development focuses on orthology prediction and mapping of next-generation sequencing data, including circular fragments.
The methods are used for instance, to build high-resolution phylogenies or to compare and re-annotate certain clades with exceptional properties.
Examples are the Aquificales, a thermophile order of bacteria or
the genus of Enterococcus that is persistently found in food such as dairy and meat but also has members well known for their pathogenicity.
The pharmaceutical focus lies on host-pathogen interactions, where we study the mechanisms of viral infections.
More specifically, we aim for a deeper understanding into the regulation of transcription and replication of the Ebola virus and
analyze the stability and influence on pathogenicity of e.g. the Influenza NS1 protein or the plant virus protein p19 that suppresses the RNA interference mechanism of the host.
In detailed evolutionary and functional studies we investigate 6S RNA and Ribonuclease P.
6S RNA is a small non-coding RNA found in bacteria.
Despite being an RNA template, it mimics an open DNA promoter and is thus able to influence transcription on regular promoters.
The molecule acts as a kind of storing device for RNA polymerase when nutrition is sparse.
In contrast to most bacteria, Firmicutes express an additional variant of 6S RNA. Its distinctive function is so far unknown.
Ribonuclease P on the other hand is found in all domains of life but can consist of a single RNA, a single protein to a variety of subunits combining both.
It is mandatory to cleave primary tRNA transcripts at their 5'-end. Nevertheless, no counterpart is known for Aquificaceae so far.
Moreover, a variety of non-tRNA substrates was discovered, suggesting a more general function in RNA metabolism.
- RNA sequencing (RNA-seq)
In the age of next-generation sequencing, RNA-seq data is generated in huge amounts on a regular basis.
It provides a detailed picture of RNA transcripts present in cells that can be looked at from a variety of perspectives.
Typical analysis in our lab cover the determination of transcription levels, the identification of transcription start and termination sites,
analyses of small and circular RNAs, RNA processing patterns (e.g. tRNA-processing, RNases) as well the identification of RNA binding sites of certain proteins (via iCLIP).
At the algorithmic side, we aim to develop methods for specialized tasks, e.g. to capture RNA processing patterns and circular transcripts
and we aim to improve the mapping of RNA-seq data, e.g. by integrating methods taken from digital signal processing.
- Orthology and the evolution of genes and gene families
Orthologous genes arose via a speciation event rather than a duplication event.
In other words, these are most likely the same genes in different species providing implications to their function and evolutionary history.
We use this data to determine genes that are specific to a certain group or special to another,
predict pathogen-specific drug targets or differences in metabolic capabilities but also horizonal gene transfer.
With Proteinortho we developed a highly parallelized, widely used orthology prediction tool which is improved and maintained continuously.
Here we aim to improve the prediction speed to a sub-exponential runtime and to further raise the specificity, e.g. by combining similarity data with gene position data (synteny) or sub-sequences of high information context (protein domains).
- Comparative genomics
Looking at the genome of a single species does not necessarily make it possible to derive its fundamental features.
Taking related species into account, allows us to distinguish between conserved and divergent elements which reveal functionally equivalent features and specific traits.
Combined data e.g. gives rise to genes that were unknown so far as well as to regulatory elements at a genomic level.
It is also used to trace genomic rearrangements and horizonal gene transfers.
The field of comparative genomics actually represents a typical application of orthology analysis which can be used as a kind of anchor the general assignment of respective loci in a set of species or strains.
- RNA binding targets, secondary and tertiary structure
RNAs act as information carriers and regulatory elements in cells.
In most cases function can only be inferred by looking at their secondary structure which is conserved while the nucleic sequence itself often is not.
Secondary structure data can be used to estimate the accessability of RNAs to proteins or compounds as well as to predict the efficency of RNA-RNA binding, e.g. to identify targets for smallRNAs or to design RNAs for anti-sense based therapies
for cancer or viral infections or to knock down transcripts in general.
- Structural biology and virtual molecular dynamics simulations
3D models of proteins and non-coding RNAs provide the opportunity to analyze folding and refolding, protein-protein as well as protein-RNA interactions and the flexibility of molecules
at an atomic level. The effects of mutations to these molecules can be predicted in silico prior to any laboratory experiment
and chemical interaction can be monitored in great detail.
Although these simulations are computational expensive, the MaRC2 computer cluster (Marburger RechenCluster 2) enables us to perform them at the range of several hundred nanoseconds.
Findings can be complemented with results of various in vitro imaging techniques such as atomic force microscopy.