Termin

Titel:
Bio-Data Science meeting

Datum / Uhrzeit:
16.05.18   /  17:00 - 20:00

Veranstalter:
Kooperation zwischen Medizinische u. Mathematisch-Naturwissenschaftliche Fakultäten (Prof. Dr. A. Dilthey u. Prof. Dr. G. Klau)




Beschreibung:

Tobias Marschall (U Saarbrücken & MPI Informatics):

Data Science meets Long-Read Sequencing
Emerging single-molecule sequencing technologies deliver long reads that can span tens or even hundreds of kilobases. These extreme read lengths allow us to interrogate genomic loci inaccessible to short reads, among which are many clinically relevant loci. Long reads enable previously impossible tasks in de novo genome assembly, structural variation calling, and haplotype phasing, but the very high error rates entail significant algorithmic and statistical challenges. In my talk, I will present three applications where data science techniques help us to overcome these challenges: (1) We show how to perfom data integration across multiple genomic technologies to achieve whole-chromosome haplotype phasing, and I will present comprehensive technology comparisons we did in the frame of the Human Genome Structural Variation Consortium (HGSVC); (2) we demonstrate how to leverage noisy long reads for genotyping, an application for which (surprisingly) few computational methods exist so far; and (3) we show the power of single-cell template strand sequencing (Strand-seq) for clustering long reads by chromosome prior to assembly, a technique that can significantly simplify de novo genome assembly, avoid misassemblies, and has the potential to aid haplotype-resolved genome assembly in the future.

Andreas Weber & Alisandra Denton, Institute of Plant Biochemistry:

Using large-scale data to discover novel biology – application to photosynthesis research
Life on earth is powered by photosynthesis. Plants, algae, and photosynthetic bacteria convert light energy into chemical energy and thereby constitute the base of most food chains on our planet. We will report on how large-scale data can be used to gain novel mechanistic insights into the inner workings of photosynthesis. The journey will start from ancient, 20th century technology, such as microarrays, via more recent NGS approaches, towards predictive biology, where we attempt to use Deep Neural Networks to capture and utilize information from DNA to predict gene expression and gene expression networks. We are trying to open the black box, and understand what deep networks see in biological data.


Verantwortlich für den Inhalt: E-Mail sendenStabsstelle Presse und Kommunikation