With its base at the Computer Faculty of the University of the Basque Country (UPV/EHU), the Intelligent Systems Team is working on various projects related to the use and application of algorithms. For example, they are helping in estimating and predicting the biomass of anchovies, constructing descriptive models through data analysis. They can also establish the most efficient route to follow for urban waste collection trucks, through techniques based on optimisation. And they can shorten the response time for computers using various calculation tasks, by means of what is known as High Performance Computing.
Algorithms can even help to better understand certain diseases, as well as to find biomarkers related to their diagnosis and prognosis. This is one of the key functions of bioinformatics. In fact, four members of the Intelligent Systems Team (José Antonio Lozano, who is the director of the team, and Borja Calvo, Iñaki Inza and Rubén Armañanzas, the latter currently at the Polytechnic University of Madrid), are working closely with researchers from Biodonostia, the first health research body within the Autonomous Community of the Basque Country and the premises of which were inaugurated this past November. According to Mr Inza, “when they (Biodonostia) see a problem in identifying biomarkers, as they are dealing with a diagnostic-prognostic model on the basis of a great mass of initial data, they come to us”.
It is researchers in the neurosciences area at Biodonostia led by Doctor Adolfo López de Muniain that have expressed the greatest interest in the tools developed by these informatics engineers from the UPV/EHU. They have worked together on studies linked to Parkinson’s, frontotemporal dementia and muscular dystrophy. Nevertheless, as Mr Inza explained, “at present the most visible fruit of our work is with multiple sclerosis. We have published an article in an international journal (in the US Public Library of Science, PLoS ONE, in 2009) and there is a patent pending between Osakidetza (the Basque National Health Service) and the UPV/EHU”.
Mr Inza stated that “whoever finds the biomarkers for multiple sclerosis will receive a Nobel Prize”, to underline the difficulty of the challenge. But at least they believe that are taking steps in the right direction. In the words of Mr Borja Calvo, the Biodonostia researchers suspect that some of the molecules known as micro RNA could be linked to multiple sclerosis, or act as biomarkers, which is why they have taken samples and analysed their levels of expression. This was when bioinformatics came into play: “They generated these data, they passed them on to us and we aimed to construct a classificatory model which, introducing levels of expression into it, was able to predict if there was a disease or not, or the state thereof”. The results were quite good: “The models predicted the disease quite well and, on this basis, a series of validation phases has been initiated”.
And how are genes converted into numbers? One of the keys to this are DNA chips; a device whereby, while fitting into the palm of the hand, stores all the known genes of the human being in a synthesised manner. “When the DNA from a sample of a person’s body is inserted behind the chip, each gene goes to its allotted slot, as it were”, stated Mr Inza. Then images of colours, partitioned into these slots, are obtained. These colours represent “levels of intensity and are proportional to the level of expression of each one of these genes. These are translated into numbers”. José Antonio Lozano adds, “The numbers express a level of fluorescence, the intensity of the signal”. With these numbers, computer models enter the scene.
Changes in the manner of research
Less than a decade ago and coinciding with the surge in bioinformatics, DNA chips started to be commonly used, Mr Inza explained. The human genome project gave a great boost to this discipline which, he stated, is partly changing the way in which biologists and doctors work and carry out their research: “A great number of techniques for recording biological data on a massive scale were developed. In the same time that a biologist used to take to obtain a level of expression of one gene, now the levels of practically all known genes can be extrapolated”.
Thus, thanks to bioinformatics, researchers no longer spend years on one single gene, and this has changed their way of undertaking research. Mr Calvo explained that “now, although you might have a concrete hypothesis, the data gathering is very general. Before bioinformatics, the hypothesis was that it was this gene in concrete involved. Now, the hypothesis is that there is some gene involved. Before bioinformatics, it was much more difficult to arrive at the goal, to find a single gene”. Now, however, what is involved is searching amongst an immense haystack until you come up with the needle. “This is where we come in – to weed out the haystack”, concluded Mr Inza.English translation by: WORDLAN firstname.lastname@example.org; 615740862.