Here is where I talk about the current status of my PhD. If you have any queries or collaboration ideas do not hesitate to contact me.
I am currently working on two aspects with a common downstream cause. Each aspect further subdivides into areas of investigation which have been pursued to various extents.
In summary, I am currently engaged in microfabrication (rapid prototyping), molecular biology (cloning, gene assembly), optimisation (MatLab, gene assembly design) and development (cell holder).
I work within the sensor research laboratory (SRL) of the Engineering department under James Covington. Here I have access to a customised EnvisionTec Mini Micro Stereo Lithography (MSL) rapid prototyper. This machine allows the design and rapid manufacture of parts for use in my project. The process in brief: The material is a liquid acrylate resin (R11) which cures to solid when exposed to UV light. Parts are designed using 3D modelling software (SolidWorks). The 3D part is then 'chopped' into 2D slices each 25 microns thick, thus a 5 mm thick part will consist of 200 layers. A Z-stage drives a glass build platform into a tray containing the resin until there is a 25 micron thick layer of resin between the base of the glass build platform and the bottom of the resin tray. The first slice is projected through the 25 micron thick resin layer, curing it in the pattern of the slice. Practically any pattern can be projected, depending on some design constraints. The resin tray they peels away from the build platform, leaving the layer part attached to the build platform. The Z-stage then drives the nascent part into the resin until the bottom of the previous layer is 25 microns from the bottom of the resin tray, another slice is projected and a layer cured. The process repeats until an entire part has been made.
The machine has been upgraded so as to maximise the resolution: The projector has been raised and focus adjusted for small parts. Additionally an ERM has been fitted which is supposed to futher increase the resolution of the device. The minimum feature size in the x-y plane is on the order of 100 microns and 25 microns in the z dimension.
Recently, following the addition of a trained chemist (Simon Leigh) to the lab novel resins have been developed which promise to expand the range of devices which could be made.
I am performing a discretised function minimisation. I am interested in finding an efficient method of exploring the state space. This problem is associated with effective gene assembly from chemically synthesised DNA.
Motivation: Imagine a set of DNA sequences, all largely (90-95%) identical, with specific mutations in. A site can be either mutant or wild-type (WT), the rest of the sequence is unchanging. The sequences represent a Hamiltonian Cycle/Gray Code where every possible combination of a mutations (0 for WT or 1 for Mutant) is covered, thus there are 2^n sequences in a set where n is the number of mutation sites. Now consider a set of discrete variables (say 10) which represent positions along this sequence. The sequence between adjacent positions is associated with a specific melting point. I wish to minimise the range of melting points for all the sequences in the set. In other words, I have 10 discrete variables (two fixed to the start and end of the sequence) and on the order of 512 sequences.
I have so far investigated only deterministic solutions to this problem.
Molecular biology (Gene assembly)
I am attempting to assemble sequences designed in part by the above mentioned optimisation process into gene-regulatory regions with Georgy Koentges. This assembly process has been used successfully in the past to assemble coding regions for genes. We wish to assemble libraries of combinational mutants of gene regulatory regions which we suspect to be integral in the regulation of MyoD. MyoD (Myogenic determinant) master muscle regulator of myogenesis (the process of muscle cell specificiation in the early embryo) and is conserved since early vertebrates (100 million years). It is hoped that by assembly of massive combinational libraries of mutants that many mutations can be investigated simultaneously, vastly increasing the rate at which such studies are done.
Development (Cell holder)
A widely applicable device which is used in conjuction with the libraries of combinational mutants decribed above. The basic aim is to remove extrinsic noise from fluorescent intensity time course data. Removing extrinsic noise is critical if one hopes to observe what might be slight changes in overall gene expression. The source of extrinsic noise is that measurements of the affects of stuff (drugs, mutations etc etc) of a reporter (fluorescent protein expression etc) over time are performed on large groups or entirely different sets of cells. Using large groups averages out expression values resulting in guassian or binomial distributions familiar to many biologists working with gene expression systems. These distributions are really representative of discretised modal changes (on or off) computed at the level of DNA itself. The problem is additionally confounded by the fact that large population of cells might contain cells are different stages of differentiation which might markedly affect their reaction to different stimuli. Currently, there is no satisfactory method for following single cells in culture over time. This device hopes to change that.
This device is simple and reuseable and is the size of a standard 384-well plate familiar to cell biologists. The device is designed to be used with a Cellomics humidified, motorised fluorescent microscope which is capable of performing many measurements over a number of days. The growth area for observed cells is restricted to a 500 micron diameter glass surface. The diameter of the observable area is the same as that of the Cellomics microscope. Cell tracking algorithms developed by Mike Downey (MOAC 1st year PhD) are envisaged to be used to improve tracking of cells within the observed area. The fact that cells are unable to move out of the observable area means that cell tracking is greatly simplified. Cells are in contact with a standard 384-well plate well volume of media which contains additional cells so that the observed cells do not suffer from feeling lonely. Individual wells are kept separate from each other so that different DNA can be applied to each well and measurements of the affects of DNA be performed simultaneously.