Speaker: Andreas von Deimling
Subject: Methylation analysis and NGS as basis for classification and grading of brain tumours.
Location: Erasmus MC
After a short introduction, the presentation by Andreas vom Deimling began. He started by explaining a little bit about the classification of tumours, and how we can group them based on gene groups. He used Medulloblastoma as an example, but he explained that this type of classification works for many types of gene groups. He went on to describe that molecular subgroups remain stable at relapse. Therefore, early tumours have the same rough pattern as late tumours. The next part of the seminar was about mechanisms that can be used to determine what kind of tumour you are dealing with. Firstly, a little bit was said about a so-called ‘decision tree’. A decision tree works with a ‘majority vote’ system. Another method is to look at copy number variations.
The speaker went on to look at diagnostic differences. He described that in 25% of cases, a neuropathologist comes to a different conclusion than a classifier. Additionally, he showed us a few reports and explained the process of classification. Moreover, he showed us the diagnostic pipeline, which describes the process of diagnosis. The average time that this pipeline takes is 10 days. After methylation profiling, there could be a change in diagnosis. Meaning that the diagnosis of the neuropathologist is changed and therefore the treatment method is changed. Towards the end of the presentation, the speaker mentioned the 450k (850k/EPIC) array technique. This technique is applied in a number of cases such as:
* The diagnosis is unexpected
* The course of action is unclear
This image shows the ‘profile’ of tumours. Such as where they are located. Diagrams like this are often used in tumour classification.
I found this seminar much easier to follow than the last one. I still learned a lot, but this time I was introduced to the topic a little bit more by other Nanobiology classes. I found the topic very interesting and especially very useful. The speaker actually introduced a website which could be used to analyse tumour data called www.molecularneuropathology.org. I found it very intriguing that they can already make information like that available to everyone and I definitely think that this type of tumour classification should be developed further in the future.