Search PubMed⌕ Search

Biomedical subjects

W Bains

Publications and source records attributed to W Bains.

At least 19 recordsLinked to original sources
PubMed2002

Implications of the network structure of cellular control circuits for the design of anti-cancer vaccine therapies.

I describe the use of a statistical mechanical model of a heirarchical network of failure-prone elements as a model for the failure of a network of interconnected biological elements that are or could be involved in the causation of cancer. From its topology, such a model predicts some of the properties of observed oncogene/tumour suppressor gene networks. I show that the same model suggests that a cancer vaccine strategy that targets 'key' network elements, such as p53 or RAS gene products, is not the only effective strategy for a broadly effective anti-cancer therapy. A mixed antigen or whole cell strategy could be as effective as single antigen vaccines, even if the latter are targeted at the most commonly mutated gene products.

Antigens, Neoplasm↗
PubMed2000

Statistical mechanic prediction of non-Gomperzian ageing in extremely aged populations.

Observations of large populations of ageing organisms have suggested that the exponential increase in the rate of age-related mortality (Gompertz rule) declines in extreme old age, and that the mortality of the extremely long-lived is independent of age. I show that this phenomenon is an expected outcome of ageing being the effect of stochastic failure of highly interconnected networks of elements with constant failure rates, and does not require the postulation of a separate 'long-lived' population of individuals.

Aged↗
PubMed1996

Family likeness.

Explore the source record for details and available documents.

Adult↗
PubMed1996

Company strategies for using bioinformatics.

Bioinformatics enables biotechnology companies to access and analyse their growing databases of experimental results, and to exploit public data from genome programmes and other sources. Traditionally occupying the domain of a 'guru' supplying answers to infrequent research questions, corporate bioinformatics is breaking down under the flood of data. New, more robust, professional and expandable systems will give scientists effective access to new tools. This review outlines how companies have evolved beyond the 'guru', and have organized their bioinformatics by acquiring or developing bioinformatics resources. It also describes why the biologist must be central to this process, and why this is a problem for computer professionals to solve, not for 'gurus'.

Biotechnology↗
PubMed1994

A spectroscopically interrogated flow-through type toxicity biosensor.

A simple immobilised cell biosensor is described which uses the MUVS (Metabolic Ultraviolet Spectroscopy) response as an assay for toxic compounds in water. Bacterial cells are immobilised in an agarose membrane in a flow cell. A sample of water is passed over the flow cell. The UV absorbence of the membrane, compared to a similar membrane lacking bacterial cells, is measured at 200 nm. When the cells are stressed so that their energy metabolism is reduced, their UV spectra change, and specifically their OD200 drops within 15 seconds of exposure to the toxin. The sensor has potential for use in rapid tests, and possibly continuous monitoring, of environmental samples for toxic contamination.

Biosensing Techniques↗
PubMed1994

Selection of oligonucleotide probes and experimental conditions for multiplex hybridization experiments.

Different DNA probes hybridize under different conditions. I examine the constraints of the design of oligonucleotide probes that are meant to hybridize to different unique sites in human genomic DNA under a single set of hybridization conditions as a parallel array. In 522 kb of human genomic DNA, 75% of 12-base and 89% of 22-base are unique, as opposed to 90% and 100% as expected of unstructured DNA, and this is not due solely to repetitive elements in the DNA. Hybridization in TMAC to reduce A+T content effects on melting temperature allows only 90% of unique targets to be hybridized under one set of conditions if a 2 degrees C difference between matched and mismatched sequences is required. Standard hybridization conditions allow no more than 60% of unique probes to be used together. This suggests that probe, hybridization conditions, and instrument design for multiple-probe hybridization applications will be harder than previously suggested.

Base Composition↗