We live in a post-genomic era. The sheer volume of data generated by genome sequencing is unimaginably immense, but all that data is useless if all it does is languish in a database somewhere. The computing capacity required to crunch through miles of sequencing data is hard to come by, especially for analysis that involves the brute-force testing of every single possible combination of hundreds of datasets. A novel approach sidesteps this consideration – instead of a single computer doing the work, the processing power of tens of thousands of idle smartphones is harnessed.
The Vodafone Foundation, a registered charity in England and Wales, is the key beneficiary behind Dreamlab. Starting in Australia, and moving now to the UK, they have partnered with academics in research institutions to establish large-scale analyses of cancer research data. In the UK they are working with Imperial College London on Project DRUGS (Drug Repositioning Using Grids of Smartphones), which is trialling multiple drug combinations against millions of cancer samples. The aim of Project DRUGS is to find which drug combinations work best against cancers with different genetic mutations. The researchers hope to make better use of currently available anti-cancer drugs, either by using them in novel combinations or against cancer types they are not usually used in.
Instead of classifying the cancer samples by location (e.g. lung or breast cancer) and stage (how far the cancer has spread), the samples are classified based on the genetic mutations that are making them cancerous in the first place. These tend to be mutations in genes controlling growth and DNA repair. While two patients with the same cancer may share some key mutations, they will also have some which differ. Similarly, two patients with very different cancers may share some mutations. Determining which mutations make cancer cells sensitive to a drug means doctors will be able to prescribe patients more effective, less toxic drugs based on their personal data.
For example, lapatinib is a drug commonly used against breast cancer tumours. It acts by targeting cells with mutations in a gene called EGFR, which usually controls cell division. Mutation of EFGR causes out-of-control signalling activity by the receptor which it codes for, and results in uncontrollable cell proliferation. A patient with a breast tumour who has this mutation will likely respond well to chemotherapy using lapatinib, as the drug will stop cell proliferation and stop the tumour from growing. However, another patient with a similar tumour caused by a different mutation will not show any response and will continue to get sicker.
Last year the FDA approved the first drug to ever be specifically developed to work against a specific cancer mutation, as opposed to a cancer in a specific tissue. However, there are other drugs out there which are currently only approved for one or two cancer types but could have revolutionary effects in some patients with different cancers. Additionally, Project DRUGS is examining some drugs which have never been used in cancer before but may still have anti-cancer effects when used in combination with other drugs. The hope is that in the future, personalised cancer therapies will become the norm.
The process of gathering this genetic data from so many cancer samples produces huge amounts of data - then data from multiple drugs in all sorts of combinations needs to be factored in. The datasets produced would take a normal computer around 300 years to process without breaks. Instead, Dreamlab uses multiple small computers – smartphones – to achieve the same result in a predicted three months.
First, the large datasets are split up into smaller packets using an algorithm. Once the app is downloaded, and the user’s phone is charging and above 80 per cent battery (to prevent power drain), two packets of data are downloaded and the processor in the phone is used to perform multiple complicated calculations. Then the outcomes are returned to the Dreamlab server and another two packets of data are downloaded for analysis. It’s a simple but clever approach, and the more phones that participate the faster the data will be processed.
The app has a clean, easy-to-navigate interface with links to information about the projects and a screen which allows you to track your personal contribution. If you are a Vodafone UK customer, this won’t eat into your data allowance, but for other network users it will. Once downloaded, all you have to do is plug your phone in at night and let the data crunching begin (this happens automatically for Android users but Apple users will need to manually activate the app.)
Dreamlab is a great way to contribute something to the scientific community without spending a penny or even putting in any work. Time to stop feeling guilty over those naps and start helping to cure cancer!
Dreamlab is available to download for free from the Google Play Store or the App Store.