A Raspberry Pi Gel Imager for Less that 100

Sunday, April 5th 2015

Print this page

Dr Ian Pringle, University of Oxford.

In a molecular biology lab one thing we have to do a lot is to analyse images of DNA on agarose gels. Its essentially a trivial matter, expose an appropriate gel to UV light and DNA fluoresces, then take a picture electronically and store it.

Equipment that does this should be quite cheap but sadly they are not as even a basic system starts around £2000+. 

We had a very good system that served us well for many years but a couple of years ago the electronics failed and we had to dig out some old equipment to get a usable system. It was so old that it stored a maximum of three images on floppy discs which magically we still had an unopened box of.

This worked - but in a lab where very few computers have floppy drives it was becoming very tedious. The PhD students were getting stroppy!


Our old gel imager complete with floppy drive storage

So what to do, we could have bought a new system but there is just something wrong in spending a lot of money on something that should be very cheap, especially when much of the actual hardware for our old system (the lense, the UV source and the darkroom) was still working perfectly well! We take decisions on spending money very seriously, sometimes its easy to justify spending large amounts of money but this just did not sit well.

Enter the Raspberry Pi and its nifty little CCD camera. Inspired by Dave Cox's lab at the University of Surrey who had used the Pi to take images from a microcope we set out to build our own Rasperry Pi imager out of parts of our old system and some free software (Pi Vision - no command line, runs in the standard Raspbian GUI).

The Pi is so small and versatile that it is really ideal for this task. First we took an old CCD module from the imager and removed its internals as it was still needed to mount the lenses. Then with a little bit of bravery we cut holes in the Pi Camera board to fit into the CCD module and removed its lense so the CCD could focus through our existing optics. Then it was simply a case of attaching the whole thing together and installing the software on the Pi.


CCD housing

The image quality is superb though the aperture of the standard Pi camera is a bit small. But these things are so cheap to build that as soon as there is a new one available we can upgrade it. All of this was done in a couple of hours (after a lot of thinking time) and the end result is that we are back in business and have saved well over £2000.

The Rasperry Pi represents a great resource for researchers to build and prototype new tools very cheaply and quickly. We are currently planning to use one to monitor the temperatures of our -80 freezers (someone has already done this too), some of which work perfectly well but their LCD displays have failed. We are also interested in a couple of motion detection alarm systems.


The Pi Imager from above with the Pi in clear case and the CCD board held in place with Blu Tac in order for this to really qualify as science!


The complete system, its not the tidiest but it works. We've subequently replaced that expensive monitor with a cheaper older one.


Large scale lentivirus production in suspension culture.


Proposed 3D model of the CFTR protein.


Light microscope image of a human airway liquid interface cultures. Dark patches are mucous.


Pellets of DNA following precipitation.


A pellet of E.coli containing a plasmid expressing a pink fluorescent protein.


A CFTR Western blot, to confirm protein production in cell culture.