In the first of our new series of ‘meet the researcher’ blog posts, let me introduce Philip Fowler, who leads our Tuberculosis-fighting project, BashTheBug
Researcher: Philip Fowler, Senior Researcher
Location: John Radcliffe Hospital, University of Oxford, UK
What are your main research interests?
Who else is in your project team? What are their roles?
I’m part of a project, CRyPTIC, that is collecting samples of tuberculosis around the world. So there are lots of scientists I’ve never met in labs in other countries who prepare the samples, inoculate the plates and take the photos that ultimately end up on BashTheBug.
Tell us more about the data used in your project
All the photographs you see are for a series of wells containing a single antibiotic. There are a few with 5 or 8 wells, but most antibiotics have either 6 or 7 wells. As you head from left to right, each well contains double the amount of antibiotic as the one before. Each strip of wells has been cut from a single photograph of a 96-well plate that contains 14 different anti-TB drugs in total. And each plate has been inoculated with a sample of M. tuberculosis taken from a patient somewhere in the world and incubated for two weeks. We hope to process 30,000 plates which means 6.3 million classifications over the next few years…
How do Zooniverse volunteers contribute to your research?
They help us by deciding in which wells the bacteria are growing (and therefore the antibiotic isn’t working) and which wells they don’t grow.
What have been the biggest challenges in setting up your project?
Umm, writing the software that cuts up each photo of a 96-well plate into smaller images for uploading to the Zooniverse.
What discoveries, and other outputs, has your project led to so far?
Our volunteers have found some artefacts that we completely missed. One person spotted fuzzy patches in the last well for one drug; turns out this is the drug clofazimine crystallising in the bottom of the well as it is present at such a high concentration!
BashTheBug won the Online Community Award of the inaugural NIHR Let’s Get Digital competition back in August 2017.
Having shown the project to lots of people what has been really interesting is it ends being a test of how optimistic or conservative you are; people who are the latter will say any small spot they see is bacterial growth, whereas the former tend to say “nah, that is really small so I am going to ignore it”. Fortunately we are asking enough people that this averages out!
We are currently trying to work out how to best build a consensus from all the classifications our volunteers have done; watch this space. Anything we do intend to publish we will first submit to the biorXiv so volunteers will be able to freely download our manuscript.
Oh, and someone mentioned us in a poem.
Once you’ve finished collecting data, what research questions do you hope to be able to answer?
A subversive one: is a crowd of volunteers more accurate and/or consistent that an expert?
A left-field one: which genetic mutations in M.tuberculosis confer resistance to certain antibiotics and, equally importantly, which do not?
What’s in store for your project in the future?
Beyond a lot more data? Hmm.
We’ve got a new design of 96-well plate on the horizon, also we want to pool all our data to see if using something like Zooniverse, but in a clinical setting, could be used to help deal with difficult cases in a hospital lab.
I’m also thinking of better ways to engage with the volunteers – more soon.
What are your favourite other citizen research projects and why?
I love SETI@home as it is a bonkers idea, has been going for about 20 years and was the inspiration behind the BOINC framework (https://setiathome.berkeley.edu).
What guidance would you give to other researchers considering creating a citizen research project?
Just try creating a test project, it is easier than you think!
And finally, when not at work, where are we most likely to find you?
On one of my bikes, or playing Minecraft with my two daughters.