Power to the people

MalariaSpot is an online ‘game’ that requires participants to count infected red blood cells on a digitised image of a microscope slide. It takes a little practice but can be mastered relatively easily. Gamers can then be let loose on ‘real’ samples. ‘Ordinary people can be trained to count these parasites,’ says Miguel Luengo-Oroz, an engineer at Madrid Polytechnic University in Spain, who developed the game.

Inevitably, those who may have the disease might be nervous about their blood samples being analysed by non-specialists. But the key is to combine the results from a number of players. If 10 people analyse an image, and nine of them find a parasite, there is probably a parasite. ‘We did statistical analysis,’ says Luengo-Oroz. ‘We found that the combined results from 13 people – each of them trained for one minute – gives perfect parasite counting, at the level of an expert microscopist.’

In time, he says, the technique could be automated: a mobile phone could be converted into a microscope – perhaps with a clip-on lens; a blood sample could be smeared onto a slide and a picture taken of it; the uploaded image could then be analysed by MalariaSpot gamers, and a diagnosis sent by text.

MalariaSpot could, says Luengo-Oroz, be applied to other samples that are analysed in the same way, such as pap smears for cervical cancer, or sputum smears for TB. And he hints at the kind of people who are likely to be most in skilled in the art of analysing these digitised blood samples. ‘Children are now technology natives,’ he says. ‘They are used to playing in detailed worlds that are much more complex than these medical images.’

The similarly themed BioGames app, developed at the University of California Los Angeles, US, shows pictures of individual red blood cells and asks players whether or not they are infected. The game can pool results from multiple players, and deliberately gives less ‘weight’ to results from poor players. In effect, bad gamers contribute as much to a diagnosis as good ones.

Phylo, developed at McGill University in Canada, requires players to drag coloured squares across the screen and line them up. It looks like a kind of two dimensional Rubik’s Cube, but could be a useful tool for molecular biologists. This game allows the players to identify similarities between different genomes, offering biologists a way of tracing the source of certain genetic diseases. Lining up the colours is akin to lining up portions of the different genomes – though absolutely no biological knowledge is required in order to play.

One of the most far-reaching Citizen Science ‘game’ projects is Foldit, which was developed in the US at the University of Washington. In it, players try to work out the structure of three dimensional simulated protein molecules, which is fundamental to how proteins function. The extended amino acid chain automatically folds into the most stable, ‘low energy’, form. The challenge for Foldit players is to manipulate the computer model into this ‘ideal’ structure.

Foldit grew out of a distributed computing project, Rosetta@home, which used people’s home computers to perform the folding. ‘Several volunteers, after watching Rosetta on their screensavers, thought they could do it better,’ says David Baker, a biochemist and computational biologist, who leads the study. The software was expanded to allow ‘interaction’ with the structures – and Foldit was born. Although computers can identify the ‘ideal’ structures, it is computationally intensive – and better suited to the human brain, which has the ability to visualise three dimensional structures.

Cracking the structure of a disease protein – such as one contained within a virus – can help scientists develop a therapy against it. Indeed, the structure of a protein from a retrovirus related to HIV, has eluded scientists for over a decade. Working in teams, Foldit players went through three design iterations for the protein. Their final structure – developed in just 15 days – fitted perfectly with experimental results, and has provided useful information in the design of antiretroviral drugs.

Meanwhile, researchers in Norway say that citizen engagement could provide more relevant environmental information, such as local air and water quality. These data are usually supplied by measuring stations, says Arne Berre of Sintef, who is involved in the pan-European Citi-Sense project. ‘There are very few of these stations,’ he comments, ‘which, as well as being expensive, are often in remote locations.’

Berre explains that weather stations based in the countryside will produce very accurate data, but have little significance for people living in a nearby city – where pollution is likely to be higher. ‘If we can engage citizens in collection, we will get much more relevant data,’ he says. Further, the advent of smartphones – and small, cheap sensors – means that environmental data can be collected by anyone and easily transmitted to a central point.

To date, the project has developed sensor packages for bicycles, and for carrying while walking. The idea is to set up a moving ‘network’ that provides local, relevant environmental data. Berre says anything that is detectable by sensors could be measured in this way, including UV exposure, particulates in the atmosphere, gases like carbon monoxide or ozone, and even pollen. As well as being connected to people’s smartphones (via Bluetooth), sensors might also be installed on buses and streetlamps.

Each individual ‘data collector’ can also be warned if they have been over-exposed to something, such as UV light. ‘That goes beyond traditional citizen scientist,’ says Berre. ‘Usually, you are only collecting data. Here, we can provide useful information in return.’

Berre acknowledges, however, that there are challenges to the ‘citizen scientist’ approach to data collection. Professional scientists and technicians are trained to take measurements as accurately as possible. This cannot be guaranteed with citizen scientists. At the same time, the sensors used in a weather station will be far superior to the ones used in this project.

But, he says, there are ways of overcoming the inevitable inaccuracy that comes with their use of citizen scientists. ‘There are algorithms to process information that is collected from many sources. It averages out the data and removes outliers, to give us more confidence in the results.’ The researchers will also compare the collected data with reference data, to further improve accuracy.

The use of computers and mobile phones suggests that citizen science is a new phenomenon, but this is not the case. The Victorians were passionate amateur scientists, and their observations – mainly on natural history – were published in a wide variety of contemporary magazines. Their contributions helped to shape one of the most famous works in science – Darwin’s On the origin of species.

‘Darwin did not have a university post, but he read these journals assiduously,’ says Gowan Dawson, professor of Victorian literature and culture at Leicester University, UK. ‘His home was his laboratory.’

Because Darwin recognised that his theories were counter-intuitive, he knew that he needed plenty of evidence to back up his claims. ‘For the first half [of Darwin’s book], you have fact after fact thrown at you – on everything including pigeons, chickens and domestic animals.’

He gleaned much of this information from periodicals such as the Magazine of Natural History, which were informed by ‘fact gatherers’, that is, local people – such as pigeon fanciers – who were the highest authority on their subject. ‘These working class men and women knew more about their subject than any university-educated person,’ he says.

This approach is still used today: ‘elite’ professional astronomers, for example, rely on their amateur counterparts to carry out ‘data processing’ – such as scanning through images looking for aberrations. ‘Lots of modern science is based on “big data” – and you need armies of people to process it,’ says Dawson.

At first glance, the Victorians who catalogued plant and animal species have little in common with today’s citizen scientists, hunched over computer images of star maps or malaria slides. But there are many parallels, says Dawson. For a start, both eras are defined by an ‘information revolution’. In the 19th century, the steam printing press gave rise to an explosion in magazine publishing – and the many journals in which the work of citizen scientists appeared. He believes that the modern information revolution – the Internet – has helped to close the gulf between scientists and the public, leading to a resurgence in citizen scientists.

At the same time, the motivation to be a citizen scientist remains broadly similar. In Darwin’s day, natural history was seen an example of ‘perfect design’ – and treated as an extension of people’s religious beliefs. This element may have disappeared in the modern world, but Dawson says that altruism is still a key motivation: many citizen scientists want to contribute to something bigger than themselves, such as finding a cure for cancer or malaria.