The Hyve’s colleagues were happy to return to Basel, Switzerland for the SWAT4HCLS 2023 conference (February 13-16). In the past years, the conference was held online because of the Covid-19 pandemic. Our Research Data Management team of four strong women truly enjoyed the in-person conference on data integration and semantic technologies in the biomedical field. Here is an impression of our experiences and lessons learned.
The first SWAT4HCLS conference was held in 2008. What started out as a one-day workshop on semantic web technologies in bioinformatics, has over the years grown into a four-day event with tutorials, lectures, and a hackathon.
As the sales representative, I (Melissa Fernando) was joined by The Hyve’s data modelers Elpida Kontsioti and Femke Kopmels and software engineer Ewelina Grudzien.
Ewelina presented a poster in the demo and data section and gave a one minute flash presentation on Fairspace project FNS-Cloud. She highlighted The Hyve’s business focus, key benefits of Fairspace and she encouraged the audience to learn about the FNS-Cloud use case: Managing food and microbiome data studies using Fairspace.
Keynotes from the Semantics Community
We thought that the scientific programme was well planned with four keynote speakers from both the biomedical field and the semantics community.
The opening keynote was given by Katrin Crameri, Director of Personalized Health Informatics at the IB Swiss Institute of Bioinformatics. She gave a presentation on The Swiss Personalized Health Network, FAIRification of health data and key learnings from this project. The next keynote was given by Ora Lassilo, Principal Investigator at Amazon Neptune. He lectured on Reimaging/Reinterpreting the semantic web.
On the second day of the conference, Chris Mungall, Head of the Biosystems Data Science Department, Environmental Genomics and Systems Biology Division of Lawrence Berkeley National Laboratory (USA), gave an insightful keynote on Scaling up semantics: lessons learned across the life sciences. The final keynote was given by Frank Van Hamelen, Professor of Knowledge Representation at the Vrije Universiteit Amsterdam. He gave us insight into the combination of Knowledge reasoning and Machine learning. He considers this combination as the next step in semantics.
Here’s what we thought were the main takeaways from the conference
Elpida: A highlight of the conference was the insightful keynote given by Katrin Crameri. She described how the Swiss Personalized Health Network (SPHN) helps medical research institutions in Switzerland make their health-related data shareable and interoperable using semantic technologies. She further explained the challenges to combine datasets from multiple national initiatives for analysis to support research activities that improve the quality of healthcare. Another enlightening part of the conference was the industry session, with multiple excellent presentations by industry leaders from Pharma and Life Sciences who continually strive to apply semantic tools and implement the FAIR principles in their organizations. I particularly enjoyed the presentation delivered by Saritha Kuriakose (Novo Nordisk). She explained the importance of establishing a FAIR culture within an organization with a vision to convert data into an asset.
The conference tutorials provided me with an opportunity to explore novel methods and tools, such as the enhancement of Clinical Decision Support platforms using clinical rules that consider semantic frameworks.
Femke: In our field, collaboration is key. Not only in sharing knowledge and data, but also with regard to technologies. As keynote speaker Frank van Harmelen described, statistical AI and symbolic AI are often kept in two separate towers. However, great discoveries can be made when combining learning and reasoning systems. The potential of such a hybrid approach within the biomedical domain was highlighted in numerous talks during this conference. For example, we heard how machine learning can be used in combination with data models to enhance disease prediction. Arif Yilmaz, a researcher from Maastricht University presented alternative approaches on generating knowledge graph based explanations for drug repurposing prediction. PhD student Özge Erten also from Maastricht University presented her findings on predicting missing annotations in the Gene Ontology. She discovered that combining knowledge graph embeddings with the true path rule enables better prediction of incomplete gene annotations.
Attending the presentations, demos and poster sessions and talking to other experts in the semantic field gave us valuable insights into how The Hyve’s services and products can truly benefit our clients.
Ewelina: The conference was a great opportunity to get to know the community around semantic web technologies and related tools. It was a pleasure to meet and listen to all those inspiring people. They were eager to share the methodologies they applied in their work so that others can benefit from these. It was clear that a lot of people attending the conference see the importance and long-term benefits of being a "FAIR researcher" and making biomedical and health data FAIR.
This was also a nice opportunity to get more insight in the semantic technologies and standards that are being used. We could learn not only about the newest technologies and trends, but also about how this journey started and how it continues to evolve.
Having this context in mind will definitely help us make more informed decisions about the technologies we use in our daily work at The Hyve and applying these to our own tools. I believe that being a part of this community is important and we should definitely come back next year.
In conclusion, SWAT4HCLS is a great venue for anyone interested in semantic technologies and an inspiration for anyone who wants to learn how advances in this field can be applied to better support customers in Pharma and Life Sciences. We are looking forward to next year’s conference in Leiden, The Netherlands. (Practically a home match for us Hyvers.)