"
Team:Cambridge/Instrumentation
From 2012.igem.org
Contents |
Instrumentation
The aim of the instrumentation project was to develop a cheap, reliable, open-source platform for detecting two emission spectra from our biological samples. In addition to this, our design was to be portable, have full functionality in the field and support any biological sensing input desired (if the biology for this exists that is). To achieve this our design comprises 3 main parts: a mechanical chassis, an electronic interface and the supporting software. We have developed a prototype as an example of how a practical solution could be implemented, however, we realise the potential for many different realisations of appropriate instrumentation and explore some ideas with respect to this below...
Mechanical Design
The Prototype
The prototype chassis consists of 6 slots in a rotary platform that can spin about its central axis. The main reason for selecting this rotary design was cost as it was assumed that one sensor and a motor would be cheaper than having separate sensors for each sample being tested. Furthermore, it means that the device can easily be scaled in the future to meet particular needs. Each slot fits a small cuvette and housing, with reflective film (to concentrate light output) on the inside, into which the appropriately engineered B. Subtilis is placed in desiccated form. By adding the analyte and potentially some germination medium, the B. Subtilis will begin to produce light that can be detected by the sensors above. A further potential addition to the whole design is to include a thermo-regulated box which would facilitate the germination of B.Subtilis spores (A prototype of which has been developed in the Cambridge MET department as part of a separate project); the germination time could also be improved using the motor controlled chassis to perform a shake cycle.
Further investigations
Many potential mechanical designs were considered during the course of this project and the final implementation of our prototype is not necessarily our view of a "Best Solution" but also one that we could demonstrate and develop easily over the course of the summer. Some major designs investigated were:
- An integrated 2D array of detectors on a flat chip board that can be directly interfaced with an arduino. The analyte would be placed in micro-wells above the sensors and sealed off for complete darkness. Advantages:
- Compact
- Potential for better sensing accuracy
- Easier to interface with arduino
- No movement for an incubation cycle
- Potentially expensive
- Microfluidic architecture for cleaning out non-disposable containers and delivery of desiccated culture. There could be an array of sensors or a single rotary system Advantages:
- Closed system - better for transport and storage
- Fully integratable system with lower margin for human error
- Expensive microfluidics!
- Hard to debug spurious results
Electronics
Prototype
The system, as illustrated (right), is made up of a sensor, a motor, an arduino and a breadboard. A Bluetooth modem (Bluesmirf) can also act as an addition to allow the communication with mobile devices (e.g. Android tablets). The sensor is made up of two light dependent resistors that sit in a potential divider circuit allowing for a direct ratiometric output without further calculations needing to be made in the software. To distinguish between the two spectra from our bioluminescent output, plastic filter gels (like the ones used in theatres) have been glued to the top of the sensors. There is also a circuit for running a DC motor, which will turn the rotating parts of our device. The whole system has been incorporated into the mechanical chassis for an attractive, ergonomic overall design.
Further Investigations
In designing the electronic interface for our instrumentation we first considered different ways of obtaining information from the bacteria and the range of possibilities that exist for the processing of this information, below are a few important ideas we considered:
- Colour sensors Advantages:
- Very good for giving an exact value of colour - narrow band detection
- Do not provide an intensity value unless using expensive sensors that would work better on an integrated sensor chip
- More wires needed for data collection
- Use of a Raspberry Pi Advantages:
- compact portable system that interfaces easily with arduino
- cheap and highly flexible
- Requires external input and output devices
- Not as nicely packaged for field work as e.g. a tablet
- Digital readouts instead of potential divider circuit Advantages:
- No need to worry about circuit bias changing (e.g. with temperature, current)
- Processing potentially more accurate (done on a computer)
- Larger potential for noise and discrepancies between channels that are not well matched
- Less elegant solution, less intergrated
Software
Prototype
The software is being developed for both computers and android devices for extra flexibility whatever the final application of the product is. The computer software is written in the free open-source language python with cross-platform wxWidgets used to implement a GUI. The current idea is to develop a fully functional GUI for collecting sensor data from all six inputs and display these as line graphs and in a final bar chart. The data to the left was recorded by moving luciferase-producing E.coli being moved towards and away from the sensor. To the right is a screengrab of the GUI under development and testing its ratiometric ability. The higher values are when one sensor is covered, the lower for the other and no change is observed when both are illuminated equally. What is actually presented is the Morse Code of the word IGEM!!! The android GUI was made in Java using Eclipse based on Amarino open-source code. The first screen (shown below) asks for the MAC address of the Arduino device and currently the application can successfully connect and read real-time data from any Android phone/tablet (all versions are compatible).
Further investigations
Software is always the most flexible part of any instrumentation system as there are an infinite number of ways of presenting information to the data and an infinite number of ways to code this. As we are not working on a software project we tried to keep things simple, cheap (free!) and relatively light in terms of functionality. However, we did discuss other possibilities that might be considered by other programmers:
- Programming languages
- C++ Advantages:
- Extremely flexible and useful for accessing serial ports
- A huge amount of online documentation and help forums exist
- Easily expandable with wxWidgets for developing GUIs>
- Slightly cumbersome to program with
- Requires compilation
