Can the power of the public help personalise cancer treatment?

This time we’re going mobile

Tonight, leading technology experts, hackers and scientists will gather in London to get ready for a weekend with a difference – Cancer Research UK’s GameJam event. Over the next 48 hours, forty ‘hackers’ will be embedding raw anonymised gene data into a new computer game, with a working title of GeneRun.

When members of the public (aka ‘citizen scientists’) play this game, they’ll be analysing this data, and speeding up our research in the process.

The event follows the success of Cell Slider™ – an interactive website developed following a hackathon event in May 2012, which allows the public to become amateur pathologists and help speed up the analysis of our archive tumour samples.

This time, we’re going mobile – so we’re teaming up with technology titans such as Amazon Web Services, Facebook and Google to turn our gene data into an app or game that citizens can enjoy playing while on the move.

But what exactly is this data, and why do we need the power of the public to analyse it?

Copy number and cancer

Over the weekend, our hackers will be using data gathered by our scientists using a technology called gene microarrays – a powerful technique to discover key genes that cause cancer and to identify regions of our genetic make-up that are frequently faulty in different cancers.

To understand what they’re looking at, we need to have a quick look at the types of DNA errors that occur in cancer.

At a fundamental level, our genes are made up of different combinations of four chemical bases – A, C G and T – arranged into long chains of DNA ‘chromosomes’.

The genetic faults we see in cancer come in a number of guises, ranging from a change in a single chemical base (e.g. switching from A to C) to huge changes in the number of whole chromosomes or, more often, parts of chromosomes. This latter phenomenon – gain or loss of pieces of chromosomes – is known by experts as copy number alteration.

The challenge is that a lot of these copy number changes involve large regions of DNA, and cause the gain or loss of many genes (and not just the key ones that drive the cancer). In order to pinpoint the culprits, the solution is to compare the genetic make-up of tumour samples from very large groups of people and find common changes that are clearly important in driving the disease.

One way to do this is to use gene microarrays, which can analyse DNA from many thousands of samples simultaneously. But with that comes a second challenge – huge amounts of data for our scientists to trawl through.

Looking for the patterns a computer can’t

Copy number alterations are a hallmark of many cancers – but they’re a particularly dominant feature in breast cancer, where gains or losses of particular chromosome regions can help predict the course of the disease.

At the Cancer Research UK Cambridge Institute, Professor Carlos Caldas and his colleagues – whose landmark breast cancer discovery we reported here last year – have been using computer software to try and spot copy number changes.

But although the software is good, it’s nowhere near as good as the human eye for spotting subtle shifts in copy number.

Take these results from a microarray below: the pink horizontal band is, effectively, the length of a chromosome – and any peaks indicate extra copies of that particular region.

There are four regions where you can see the levels change – these show where alterations have occurred in a specific part of the genome. But to be able to home in exactly on which genes are affected, and identify those that may play a key role in cancer, our scientists need to know precisely where this shift begins and ends – and for this we need the accuracy of the human eye.

Our researchers have thousands of datasets just like this one, in many different types of cancer. In fact, researchers across the world are using these techniques in multiple cancer types as part of the International Cancer Genome Consortium, for which we’re leading the efforts to analyse oesophageal and prostate cancer.

So if we can develop a game that can help scientists pinpoint key hotspots in the genome faster, we will begin to see the fruits of this research much more quickly.

Speeding the benefits of genetic research to patients

So what does this mean for people with cancer? Professor Caldas and his team have already shown that breast cancer is actually 10 different diseases, and there’s almost certainly a similar situation in other cancer types.

The key to improving survival for all cancers is to understand its underlying genetics, so that we can give patients the right treatment, for the right cancer, at the right time.

So, by analysing this data, citizen scientists will be helping to spot key mutations in genes – and these could provide new targets for cancer drugs that could lead to breakthroughs in treating the disease in future.

As an example, painstaking work in the early 2000s showed that a gene called BRAF is faulty in the majority of melanoma skin cancers, and in a smaller proportion of other types of cancer.

Last year, the drug vemurafenib was approved to treat melanoma patients whose tumours contain faulty BRAF genes. And the drug is being tested against other cancer types when they bear this mutation.

But from start to finish, this took over a decade to do. We need to act faster.

This weekend is the first exciting step in analysing this data faster. By Sunday, we hope to have the beginnings of a game that anyone can and will want to play, harnessing the power of the public to help our scientists beat cancer sooner.

We’ll keep you posted on the results – and follow us at #CRUKGame to see our progress during the weekend.