Friday 1 November 2013

Health Hack Melbourne 2013

On the weekend I attended the inaugural Health Hack Melbourne weekend.

I went along hoping to help further medical research in a meaningful way, and to also learn something new (as a bonus).

The event started on Friday evening with mini presentations from each of the problem owners (medical researchers) that was a sell job on how solving their problem would help the world, and also gave the attendees an idea of how they can help solve that problem. The problem owners were then available for further questions/interrogation at different locations, and at the end of the night we had formed a team and started floating ideas about how to break down the problem and find the most valuable chunks to solve given the reasonably tight timeframe.

Our team page has a bit of detail about the problem for those that want to look further, but at a high level, our problem was to find a way to visualise and analyse correlations between 2 genes within a given context/dataset. The context/dataset was already an output of another process that the researcher had put together.

We worked pretty hard over the weekend as a team of 5, and ended up enabling the researcher to import new datasets, as well as filter results and enable him to identify those genes that have the highest number of relationships/correlations within a dataset. The end result was hosted on heroku (it is no longer up and running).

Our researcher was very happy with the result, not just because of what we delivered within this app, but because we introduced him to a number of new options he had for visualising data, including algorithms for determining interesting subsets of genes to focus on, as well as other out of the box visualisation tools such as gephi which could be of use.

Although we didn’t win the event, and we didn’t progress quite as far as I had hoped (we didn’t revolutionise genetic testing, or find a cure for a disease), our researcher was very happy, I met some passionate people, and I learnt a bunch of stuff I didn’t know about before the event, including just a little more about genetics/chromosomes/genes and how much data is generated out of a typical genetic test.

I’ll definitely be keen to attend next year’s event.

No comments: