Skip to main content
Premium Trial:

Request an Annual Quote

UK Charity Teams up with Google, Amazon, Others to Use Crowdsourcing to Identify Cancer Mutations


Cancer Research UK is working with the Citizen Science Alliance, Google, Amazon Web Services, Facebook, and academic partners such as Cambridge University to design and develop a free game for smartphones that will help researchers identify chromosomal rearrangements associated with cancer.

The British charity and its partners organized a weekend-long hackathon dubbed "GameJam," that was held March 1-3. They invited 40 hackers to develop prototypes for the game, whose working title is GeneRun. To that end, the organizers provided hackers with raw microarray data from breast cancer patients, which had been pre-formatted by Citizen Science Alliance, and asked them to come up with potential formats for the game.

The hackers were expected to come up with creative ways to identify "shifts" in microarray data that could indicate things like deleted genes and duplications and be linked to breast cancer, Joanna Owens, CRUK's senior science communications manager, told BioInform this week.

She explained that while computer programs that look for these changes already exist, they are good at spotting obvious genetic shifts but often miss those that are more subtle.

Human beings, on the other hand, can detect these nuances, but it would take too long for researchers to go through all their samples and locate where shifts occur on the chromosomes, she said.

GeneRun is intended to "basically open up this analysis to many, many thousands of human eyes" and get "the public involved," she said.

During the hackathon, nine teams came up with 12 possible protocols for detecting differences in data. These work in a variety of ways; for example, one game requires players to gather certain objects, while another requires players to try and kill things in response to changes in patterns that they see in the data, Owens said.

She said that CRUK will evaluate the game ideas based on criteria such as how well they are able to analyze data, how fun and addicting they are to play, and what the graphics look like. Owens also said that the final game might be based on only one of the sample games or it might combine multiple concepts.

It will then hire a third-party developer to build the game, which it expects to launch this summer, she said.

Amazon will provide the technology platform to host the final game, CRUK said. It also provided the GameJam participants with free technology resources and expertise to help them start developing their prototypes. Meanwhile, Facebook’s London-based engineering team provided expertise to the hackathon participants while Google provided an undisclosed amount of financial support and hosted the event at its campus in East London.

Although these proofs of concept were developed using data from breast cancer patients, GeneRun will be able to analyze data from all types of cancer, Owens said.

This is CRUK's second shot at using crowdsourcing to help researchers analyze oncology data. Last year, it worked with the Citizen Science Alliance to develop Cell Slider, an online game where players are shown images of stained breast tumor samples from clinical trials and are asked questions, the answers to which are used to categorize the samples.

The stain in the samples marks the presence of an estrogen receptor that is associated with patients' response to tamoxifen, Owens explained. Players' responses to the questions provide pathologists with "an idea of how strong estrogen receptor expression is in those samples, which is linked to the success of tamoxifen treatments."

CRUK launched a beta version of the Cell Slider website last October. Since then, tens of thousands of players from Citizen Science Alliance's community — including Canadian school children and a member of the British parliament — have analyzed more than 684,000 images, Owens said.

She added that CRUK estimates that if pathologists had analyzed the data themselves, it would have taken them 18 months to go through that many sample images compared to the three months it took to get results using Cell Slider.

CRUK is now checking the results of the samples analyzed by Cell Slider's players to see how they compare to samples that were analyzed by expert pathologists, Owens said. This is to make sure that the game actually works and that it returning results that are useful and accurate, she said.

They expect to launch Cell Slider fully in April, she said.

CRUK is one of several research groups currently using games to enlist the general public's help in finding answers to life science research questions.

One of the more well-known efforts is the University of Washington's Foldit, a protein folding game where players try to predict the structure of proteins. In 2011, members of Foldit's online gaming community solved the structure of a protein-cutting enzyme from an AIDS-like virus whose configuration eluded researchers for over 10 years (BI 9/23/2011).

Others include a 2010 sequence aligment game called Phylo, which was developed by a research team at McGill University. Phylo players are expected to find the best possible aligment for sequences that are represented as rows of colored squares, with one square per nucleotide, in a grid that represents the genome. Gamers can move entire sequences or individual boxes to the right or to the left in an effort to match colors in one row to the corresponding box in the row below (BI 12/3/2010).

More recently, scientists from the Scripps Research Institute developed three games — Combo, Dizeez, and GenESP — that are aimed at improving current knowledge about gene-disease links and combinations of genes associated with particular phenotypes (BI 7/20/2012).