Report from GameJam: accelerating science outside the laboratory

Tech specialists and scientists joined forces over the weekend

Imagine that you could help beat cancer just by playing a game on your smart phone. It might sound far-fetched, but we’re working to make this idea a reality.

Over the weekend of 1-3 March we brought together Cancer Research UK scientists and technology specialists to develop a mobile game to help speed up our research.

Emma Rigby from our press office watched the events unfold and shares her diary of the weekend below.

Day 1: developing ideas

It’s 6.30pm on a Friday and most people are heading home after a long week in the office but for this team of experts, the work is just beginning.

We’re at Google Campus in London for a 48 hour GameJam – an event where technology experts build computer games. As we mentioned in our last post, – this is a GameJam with a difference, bringing together Cancer Research UK’s scientists and more than 50 experts – developers, designers, gamers and mobile specialists including gurus from Amazon Web Services, Facebook and Google.

There are new ‘game jammers’ from Eventbrite and return experts from Cancer Research UK’s first hackathon event last year. There is support from Citizen Science Alliance, Omnisoft and students from Imperial and City University.

People have travelled from as far as Seattle, Dublin, Belfast and the Isle of Man to be here.

Everyone’s giving their time and expertise for free to develop a mobile phone game to speed up cancer research.

Increasingly scientists are targeting treatment to the genetic fingerprint of cancer patients. But to develop drugs for specific patient groups, scientists must know which genetic faults cause the disease.

This means analysing mindboggling amounts of genetic data. The data resembles a squashed up radio wave with sharp peaks and troughs correlating to gene behaviour in tumour cells. There are two million squiggley lines of gene data per patient – comparable with a telephone directory containing 20,000 pages.

Trained scientists scrutinize this data using computer algorithms to search for new genetic clues. But the algorithms aren’t perfect, which means they can miss key genes or mutations important in cancer.

We’re counting on the GameJam participants to solve this problem – in 48 hours.

They must design a game to detect changes in the data – indicating when there are either too many copies of certain genes – or too few.

The game must be social – so people can share it. It should have a reward system – so people play it again. And it has to be fun.

Dr Chris Lintott, chair of the Citizen Science Alliance sums up the mood. “This is fabulous but scary! We’re tackling not one but two difficult problems – doing important science AND trying to create a usable game. Luckily we’ve got the right people…”

There is a talk from Amazon Web Services and Facebooks experts, and the ice breaker is done away with – there’s no need. Everyone’s in groups cracking on straight away.

The pizzas arrive. And the 48 hour GameJam begins.

The first stage is idea generation – what is the basic idea the teams will follow up in the morning.

“What about some kind of steady hand game? What if we make the data represent 3D space – players could tilt their phone to roll balls around?” Everyone chips in and thoughts are scribbled down. “We could translate the line into colours.” “How about using an accelerometer?”

James Lau, a games producer explains: “We’re all here for the same reason. It’s the combination of tackling a fascinating challenge while knowing that what we come up with is helping a very worthy cause. It’s very exciting.”

Day 2: number crunching

Some of the participants have had three hours sleep. One girl has brought a sleeping bag.

And it’s going to be a seriously long day.

People quaff their first coffee of the morning – there are cans of energy drinks at hand for later – and the pace of work is red hot. The groups have already chosen the ideas they want to develop.

But considering each game must incorporate 40 data files, some 846 090 pieces of information, equivalent to the data on one person’s chromosomes – the solution will be anything but simple.

Six Cancer Research UK scientists mingle, helping the programmers incorporate the data into their games. “Above all,” says researcher Dr Oscar Rueda, the game must distinguish between real differences in the genetic data and background artifacts – or false results.” In other words, accuracy is key.

Lunch comes and goes. The energy drinks disappear. Techno music’s playing. Everyone’s got their heads down.

John Lau adds: “It’s going really well. There’s loads of enthusiasm and it’s really heartening to see people coming together like this. Keeping the cause in mind is really keeping us going.”

For the rest of the day it’s coffee, collaboration, designing, developing and hard, hard work. People simultaneously type codes, try out ideas and challenge team members. No one has noticed it is already dark outside.

“Did you get a sound engine running?” asks one programmer. “We need an interface to specify the start of the data points.” “I’m concerned about the distribution of data on the x axis,” says another. “It’s fine, we can mess with that.”

By 1am some have gone home but many remain. The music‘s louder. There’s 24 hours to go.

Day 3: innovation, energy and creativity

Most people have had a few hours sleep but some haven’t closed their eyes at all.

Overnight there’s been progress. Teams have developed lots of different, original approaches. Original ideas. But it’s no smooth journey.

“We’re trying to add too many graphics at once. The system is locking up!” shouts one. Another responds: “It’s OK! We’re just adding entities to the array while we’re still in the loop.” “But there’s something wrong here with this code!”

Everyone works their socks off, stopping only to grab an energy drink.

Stress levels rise. We’re at T- 2 hours. Games come together and people validate the results with scientists. But these guys are absolute perfectionists and they don’t want to finish until they have nailed it.

Researcher Dr Raza Ali, comments: “I’m amazed at the innovation, energy and creativity that has been unleashed here this weekend. I’m touched by the dedication of the individuals and how they have collaborated this weekend. This will be a major help to us as cancer researchers.”

By nightfall we see nine finished games and 12 working prototypes – of overwhelming quality.

But perhaps now comes the hardest part of all – picking the best ideas.

Over the next few days our scientists and the CSA’s Chris Lintott will thoroughly test the prototypes to see which one will generate the most accurate science, attract players – and be social at the same time.

And then we’ll recruit a gaming agency to create the game – to launch later this year.

We can’t wait to reveal the outcome.