Ignasi Belda, proteines with artificial intelligence

We are visiting el Parc Científic today and in the middle of the chaos an bustle of a neighbouring building site we met Ignasi Belda, who is reading his PhD thesis this month and opening his new Intelligent Pharma's office at Parc Científic in a few days.

Tell us how you got here...

We know how you move around during your PhD, but what is the science we find in it?

I am from Alacant and in my experience it was very positive to come to the big city. I arrived here in 1997 to study a technical degree on Computing at  La Salle School at Ramon Llull University. When I was in my third year I decided to start a research project with professor Ernest Giralt, who is now one of my PhD advisors. I worked in the application of  artificial intelligence  to the optimization of peptides. I loved computing so I decided to go on with my studies and I go into  Computing Engineering. During the following years I combined studying with the research that I had started with Ernest. In fact, he was the person who encouraged me to go on even further and start the PhD Thesis that I am reading next week.            We could say that your Thesis has many parents. How is it possible? Yes, I have worked with many institutions and I have two co-directors and a tutor. Although my situation is not the most common one, it has happened in a very natural way. I started my research under Ernest Giralt's supervision,  an organic chemistry professor from Barcelona University, so this was the beginning of my PhD. Then Parc Científic was inaugurated a little later and Ernest decided to move there, bringing me with him. In order to proceed with  my research in the field I was interested, artificial intelligent, I entered PhD masters organised by  the LSI department at UPC University where professor Àngela Nebot of the SOCO group became tutor of my Thesis. On the other hand, I also got support on the artificial intelligence world by Xavier Llorà. Xavier is a professor at University of Illinois at Urbana-Champaing and he is being the co-director of my Thesis. The Institut de Recerca Biomèdica (IRB) was born a couple of years ago and Ernest Giralt was appointed as one of the leaders of this new project. That was how I became related to IRB, too. On top of all that my stay at AstraZeneca, a worldwide leading Swedish Pharmacy company, gave the last finishing touch to my research and become the sixth institution to thank for the possibility of my project.   Well, if I am not missing anybody that would be it. These are the six PhD God fathers of my Thesis.

As I was saying before, I am an specialist in artificial intelligence applied to biomedicine. The title of my Thesis is: Evolutionary Algorithms and de Novo Peptide Design. We could say that it is structured in two parts. The majority of my research is related to evolutionary computing applied to peptide design. In this first part, I have generated evolutionary algorithms to optimize the bounding energy between some peptides and certain proteins. The biologist and the biochemist that were working in my research team needed to generate and know the kind of peptides that join nicely with these proteins. I give them the computing results through computing simulations and that helped them with the real situation. I began with the Darwinian Theory of Natural Selection and I started my job by generating a random series of peptides, or aminoacids. Then I simulated how these peptides joined the chosen surfaces. Then by using the procedure known as docking and considering the energy that peptides lose with their bonding, I was able to evaluate how these peptides adapted. I chose the best peptides from the first process and then generated a new second generation from them. The second generation was evaluated again and from the best ones I generated a third one and so on. In fact, the methodology that I use is an iterative one. On top of that, the process  also introduced some random mutations. Therefore, it follows the real life evolutionary model. This way of working is based on the use of genetic algorithms that belong to evolutionary computing. Part of my research is also directed to improve the docking process, a genetic algorithm itself. My intention is to get better quality and more speed. All this work is computingly very expensive. It is important to notice that in order to develop the evaluating part of the project we needed to work with  with Mare Nostrum and five hundred processors at the same time. On the hand, my stay in Sweden meant incorporation of artificial learning into my research project. AstraZeneca is a company with four thousand researchers and where fourteen million dollars are invested every day in research. As a pharmaceutical company, it is interested in the toxicity of certain molecules. My contribution was to develop a program that was able to distinguish between a toxic and a non-toxic molecule and to give a reason of its toxicity. It is very important both for research and production to know what toxic molecules have in common in order to avoid it when they are designing new ones. To obtain an explanation I use evolutionary algorithms with rules and parameters: electrical charge, hydrophilic ...etc. I work with more than one thousand five hundred physical and chemical parameters and I get the answer what a molecule should be like in order to be toxic. All this knowledge and the evolutionary algorithms developed amount to the last 10% of my thesis.

Intelligent Pharma is a chemoinformatisc and bioinformatics company located at the Parc Científic of Barcelona. The goals of the company are to develop highly specialized software to fulfil our customers' research requirements, and to give them computational and artificial intelligence support in their research projects based on different fields: pharmaceuticals, biomedicine...etc.

The idea of starting a company started only in the last years of my life; in fact, we could say that Intelligent Pharma was born the day after I handed in my PhD thesis, a sunny St. George day, in 2007. Nevertheless I personally believe that being an entrepreneur is something that starts when you are born; my parents were both business people and I knew I wanted to get there as soon as I started my degree.

People think that Intelligent Pharma is closely related to the job I have been doing for the last years. However it is just based on the knowledge and experience that I got during this years. During this time, I have realised, and I do not think people are aware enough of it, that there is a big demand n the health business for  people with a strong background in computing. Starting a company is a slow process; I don't want to lie to anybody about it: It needs an entrepreneurial stance, and it is hard to get enough financial background. In spite of all this  I'd love to encourage people who are thinking about  it because you can get many loans and grants. Parc Cientific, for example, has a program called bioincubators and allows newly born companies to use its material and installations. There are banks like BanCaixa that organise contests for young entrepreneurs. Intelligent Pharma won the last edition of that contest. UPC also gives support through its initiative Innova... There is  a big list of institutions willing to help science, economy and people to succeed