Researchers connected in Berlin

researchersConnected.png

I really enjoyed attending the Neo4j Life & Health Sciences Workshop, organized in Berlin, this week, by Michael and Petra: a day rich with great presentations about the application and utility of graph technology in several research areas. Here are only few examples:

  • The Ontology Lookup Service, a repository for biomedical ontologies, implemented with the support of graph databases and Apache Solr for indexing, different technologies for different purposes.
  • In the Lamond lab (University of Dundee), they model proteomics data with graph databases in order to understand protein behaviour under different conditions and dimensions of analysis.
  • MetaProteomeAnalyzer (MPA), a tool for analyzing & visualizing metaproteomics, uses Neo4j as the backend for metaproteomics data analysis software.
  • Tabloid Proteome is a database of associated protein pairs, derived from mass-spectrometry based proteomics experiments, implemented using a graphdb, which can help also to discover proteins that are connected indirectly, or may have information that you are not looking for!
  • Reactome is a pathway database which has recently migrated from MySQL to Neo4j, with relevant performance improvement. You can access data via the GraphCore open source Java library, developed  with Spring Data Neo4j, or via Neo4j browser.

I’ve lost count of how many times I heard sentences like: “Biology systems are complex and growing and graphs are the native data model” or “Graph database technology is an effective tool for modelling highly connected data as we have in biology systems”. We already knew it, but it’s been very encouraging and promising hearing it again from so many researchers and practitioners with higher experience than us in graph technologies.

In the afternoon, I attended the workshops “Data modelling with Neo4j”; starting from the data sources we usually work with, we have tried to model the entities and the relationships in order to answer some relevant questions. Modelling can be very challenging and, in some cases, it might depend on the questions you have to answer!

Before the end, I had the chance to give a short presentation about our experience with Neo4j.

Thanks again Michael and Petra for organizing such a great event!

Advertisements