The 2018 neurological phenotyping course for systems genetics – an invitation to Luxembourg

Computational phenotypes. Clinical epilepsy research requires the capturing of complex information in a way that then can be subjected to statistical analysis. For the analysis on the phenotype level, new standards are emerging that are heavily informed by genetic studies. In fact, in addition to the known domain-specific classifications such as the ILAE classification for epilepsy, interdisciplinary action is often required to improve the classification of neurological syndromes for a larger analysis. During the upcoming EMBO Practical phenotyping course in Luxembourg, we will introduce trainees in the field to concepts like the Human Phenotype Ontology (HPO), a controlled vocabulary to characterize syndromes and one of pillars of research in complex syndromes such as epilepsy and how to address aspects not covered in HPO. The course will be held in Luxembourg from Oct 4 to Oct 10, 2018. There has already been a strong interest in this course, but we have a few spots left if you would like to register!

EMBO Practical Course. This course will cover one of the most complex issues in modern biomedicine – trying to assess and transform phenotypes into a format that is accessible for a systematic analysis. We will cover the various ways that phenotyping can be made accessible for systems biology and data science. The target of this course is trainees in the biomedical area, both from the clinic and the computational side.

Why is a course for phenotyping neurological syndromes for system genetics necessary? Traditionally, phenotyping has been an issue that was exclusively performed and analyzed by clinicians with direct contact to the study participants – physicians, genetic counselors, research nurses, and other health care professionals would phenotype their patients as they simply had the most direct access to the patient cohort. One important lesson that has transpired from the recent development in human genetics is the fact that we often need larger cohort sizes if we would like to make meaningful discoveries and data is assembled from clinical records retrospectively. However, most of the domain-specific classifications are not necessarily compatible. For example, the current ILAE classification, the officially recommended classification for the epilepsies, is not necessarily compatible with classifications in other fields of medicine, such as format used in medical genetics or laboratory medicine. This makes it difficult to really explore phenotypes at depth at a certain scale.

HPO. The Human Phenotype Ontology (HPO) is currently becoming the “lingua franca” in biomedicine. It represents a vocabulary that allows for the coding of phenotypes at a certain depth while connecting all phenotypes through a tree structure that allows us to explore to which degrees phenotypes are related and how far they are apart. During our course, The HPO will be introduced by Peter Robinson, its creator. There will be active review and development of the ontology to capture aspects of neurological syndromes such as epilepsy that are not yet covered in necessary detail. For example, on the European side, the entire RES cohort is phenotyped in HPO, which allows us to compare this cohort on a phenotypic level with other cohorts such as the UK-based DDD cohort. Basically, during our course, phenotyping itself is at the core. Katie Helbig and Sarah Weckhuysen will introduce us to neurological phenotypes with a focus on epilepsy and how we get from a patient phenotype to a completed Case Report Form that can then be used for systematic studies.

Data management. Phenotypes make us revisit what good data management means. For genetic data, we have well-established formats like FASTQ, BAMs, and VCFs that we can work with, exchange, and merge into larger studies such as the recent meta-analysis by Heyne and collaborators. However, for phenotypes, we have to think about more basic issues. What constitutes good data management, how do we design our Case Report Form? During this course, I will present a segment on good data management and Kristen Roomp will discuss the ins and outs of how to create good CRFs.

Genetic data. Finally, during our seven-day course, we will also cover capabilities that structuring this data brings for genetic research. Slave Petrovski will introduce burden tests and related tests for the analysis of variants that can be heavily informed by phenotypes that have been translated into a standardized format. For example, electronic health records (EHR) provide a wealth of possible information for extraction of details for research purposes but face numerous issues starting with the availability to structure extraction. Our course will gather clinicians, statisticians, geneticists, and bioinformatics researchers to exchange ideas and concepts in the field of applying computational phenotypes.

Location. The course from Oct 4 to Oct 10, 2018 will be held in a retreat style in the Alvisse Parc Hotel, a short ride from Luxembourg city. Travel grants are available for attendees from Croatia, Czech Republic, Estonia, Greece, Hungary, Italy, Malta, Lithuania, Poland, Portugal, Slovakia, Slovenia, Spain and Turkey as well as Chile, India, Singapore and Taiwan as EMBC Member States or EMBC Associated Member States provided by EMBO. The course is sponsored primarily by the EMBO Course and Workshop program and Elixir-Luxembourg. Here is the link to register for this course if you haven’t already.

Roland Krause

Roland is a bioinformatician at the Luxembourg Centre for Systems Biomedicine. He received his undergraduate degree in biotechnological engineering and a PhD in biochemistry from the University of Heidelberg. His postdoc was in computational biology at the MPI for Molecular Genetics, Berlin, shared with the computer science and math department of the Free University Berlin.