More Information
Before the development of DNA fingerprinting, tracking of epidemics and dangerous diseases relied on conventional phenotypic identification of the pathogen, but the techniques were poor at discriminating between different strains of a micro-organism, and tended to be confined to specific groups.
DNA fingerprinting is much more powerful and has wider applicability. One particular technique, PCR-based fingerprinting, has emerged as the favoured method because it has the potential to be made reproducible, so that different laboratories will always obtain the same DNA fingerprint if presented with DNA from a given pathogen. This reproducibility can be achieved by combining tight control over the quality of the PCR reagents with the use of standardised techniques for performing the tests.
In order to identify samples gathered from the field, it is necessary to have a database of DNA fingerprints with which to compare this sample. The task of compiling this database is time consuming and requires computer assistance. Software that helps performs this task is now available, one example being the BioNumerics program developed by a participant of this Network, but there still has to be agreement over how to generate data for a database, and over protocols for data entry and retrieval. The Network will focus on this pressing task, rather than on the development of the typing methods themselves which is a problem that has largely been solved.
There remains one other fundamental problem that has not been satisfactorily solved, and that is the actual identification of the pathogens by comparing their DNA with the fingerprinting database. With existing techniques, an assumption has to be made about the characteristics of the DNA profiles being compared, and this can lead to incorrect identification. There is a potentially more reliable and efficient method based on a heuristic, or self-educating model, using a neural network structure, in which the computer "learns" the distinctive geometric features of each set of DNA. It then applies the same principle to samples that need to be identified. In this way new samples can be assigned to a set, and experience gained from other identification problems suggest that such neural network based models will produce highly accurate results.
The ESF Network is looking at the possibility of developing such a neural network model, which could eventually be used in different laboratories across Europe, all linked to a common DNA fingerprinting database via the Internet. This would allow routine on-line identification of, and comparison between, a broad range of bacteria.
The third and final task of this Network is to serve as stepping stone towards exchange and maintenance of data pertaining to DNA chips, which are sections of DNA data generated by the ongoing process of sequencing microbial genomes. Such chips are expected to become available commercially for identification of microbes by 2002, and could play a key role in the early recognition and prevention of major outbreaks of infection by virulent pathogens.
The need to detect emerging outbreaks of infection before they reach the epidemic stage is becoming even more pressing as the EU expands to include additional countries on the eastern and southern fringes of Europe. Many of these countries are bordering on regions with hotter climates where the endemic micro-organism population includes a greater number of virulent antibiotic-resistant pathogens. For this reason it would be desirable to establish "sentinel" laboratories in these countries to detect initial spread, from neighbouring regions, of micro-organisms likely to cause dangerous infections or epidemics within the EU. Such laboratories would be given access to the on-line typing databases to facilitate rapid identification of suspect pathogens.
The primary aim of the Network is to formulate collaborative multi-disciplinary proposals for developing the Internet-based database for DNA fingerprints and making it readily accessible and easy to use by microbiologists with limited computer expertise. This task has been sub-divided into nine activities, which are:
1) Establishment of a WWW Home page and electronic links.
2) Establishment and distribution of panels of well-characterised microbial strains.
3) Agreement on methodology for generating data.
4) Agreement on the optimum method and format for communicating data.
5) Establishment of an interactive Internet based DNA fingerprint database.
6) Expansion of the database by collecting data relating to virulent or epidemic microbial pathogens.
7) Development of a neural network model for automatically and accurately identifying pathogens from their DNA fingerprints.
8) Dissemination of information relating to the Networks activities, objectives and achievements to other laboratories in Europe and the rest of the world.
9) Formulation of longer term strategic recommendations, for example deciding who will be allowed to access the information, some of which will be sensitive and confidential. It will also be necessary to ensure that the database is compatible with other sources of pathogen-related data.
Laboratories or individuals interested in actively participating in the work of the ENEMTI Network can register to receive more information at the ENEMTI web-site Go to website
Activities
The network will fund:
- three workshops
- First workshop took place on 3-4 November 2000 in Bilthoven, the Netherlands
- Second workshop took place on 16-17 November 2001 in Gent, Belgium
- Third and Final Workshop (closed workshop) will be held on 22 November 2002 in Barcelona, Spain
and on Saturday 23 November 2002 a Symposium on "Latest Advances in Molecular Fingerprinting Methods for Bacteria and the Construction of Computer Databases" will be organised at the University of Barcelona.
The Symposium is open to anyone who is interested (clinicians, scientists, research technicians, students). No registration fee, but number of participants is limited. Please contact Dr. Kevin Towner to reserve your place. - website
