Interview with Florentino Fernández Riverola, PI of the Next Generation Computer Systems Group (SING)

We have the capability to take on complex problems and work in teams in which the coordination of professionals belonging to different fields of knowledge is essential

 Florentino Fdez-Riverola has a doctorate in Computer Engineering and is a member of the Department of Informatics at the University of Vigo, where he heads the Next Generation Computer Systems Group (SING), a multidisciplinary group which focuses on the research and development of new computer applications and methodologies.

What are the main lines of research for the group? What are its main added values as a research group?

The group has 6 main lines of research which are linked to knowledge representation, the use of machine learning techniques, case-based reasoning systems, the creation of hybrid Artificial Intelligence models, the use of agents and multiagent systems and the creation of decision support systems.

The fields of application of know-how held within the group are related to diverse areas such as the filtering of inappropriate content on the internet, environmental prediction, the development of new models and workflows in bioinformatics and computational biology, the IT support required for advancement in personalised and translational medicine, and also specific aspects of active and assisted living (AAL).

The fundamental value of the research group, laying to one side the excellent professional qualifications of all its members, is the capacity to take on complex problems and work in teams in which the coordination of professionals belonging to different fields of knowledge is essential.

 

Over 170 publications, almost 40 research projects and more than 50

R&D contracts; these are just a few key details about the group. What other indicators of your potential would you highlight?

One aspect that I believe highlights our collaborative disposition and our commitment to society, beyond the merits achieved in research on a global scale, is our ability to produce highly qualified professionals, and our collaborations with the free software movement and open source code.

 

One of your main areas of research is in Bioinformatics, and you have developed various applications in this field in recent years.  Which of these have been most significant?

Obviously they are all important, although clearly the function and the number of final users varies greatly from case to case. To mention a few of them, we have ALTER (http://sing-group.org/ALTER), which is a file format convertor for multiple sequence alignment. Despite seemingly having very basic functionality, this tool receives an average of 1000 visits per month, and has been cited in more than 90 research projects.

Another example comes from the field of biomedicine. InNoCBR uses a tool developed in conjunction with the Preventive Medicine Service at CHUO (Complexo Hospitalario Universitario de Ourense), which for the past year has been the work base for this team in the detection of nosocomial infections (infections originating inside a hospital).

Finally, in relation to the field of text mining I would highlight both BeCalm, a biomedical annotation meta-server, and Markyt, an open-source toolkit for the annotation, visualisation and evaluation of automatic systems dedicated to mining biomedical texts. Despite being two fairly recent projects, they are already showing very positive indicators with regard to their quality and usefulness. Currently, the integrated servers at BeCalm support the annotation of 12 biologically relevant entity types (e.g. genes, proteins, chemical elements and diseases). In only the first three months of 2017, BeCalm served 4,092,502 annotation requests with an average response time of 3.74 seconds/server. Meanwhile, Markyt has given support to various biomedical applications (e.g. research into antimicrobial resistance, etc.) and has been used in evaluation tasks promoted by the biomedical community.

 

Another area of the group with great potential is the application of hybrid Artificial intelligence models to translational research. What are their main applications?

Hybrid methods, machine learning and Artificial Intelligence in general, are very useful tools when exploring large volumes of data, as they enable the identification of characteristic patterns of determined conditions of interest (e.g. healthy patients VS ill patients). In this sense, hybrid methods offer greater flexibility by integrating different techniques to carry out the identification of these patterns.

An example of the real utility of these type of methods is the previously mentioned InNoCBR, which processes daily reports coming from microbiology (laboratory), pharmacy and nursing from all hospitalised patients at CHUO to identify the cases in which potential nosocomial infections have occurred.

In addition, we have also carried out various projects such as geneCBR (http://www.genecbr.org/) or geneCommittee (http://sing-group.org/GC), in which the use of Artificial Intelligence techniques is explored for the integration of gene expression data with biological knowledge (e.g. metabolic routes, genes associated to certain pathologies, etc.) with the aim of facilitating the diagnosis of different diseases, in particular cancer.

 

The group has an intense level of international collaborative activities, making up various networks and consortiums such as the Spanish Association of Artificial Intelligence or the Vindeira platform, for example. What is the main output of this collaboration?

Independent of the results obtained with a degree of scientific and / or economic impact, our international collaborations enable us to be part of a continuous updating process, something which is fundamental both in the technological and biomedical sectors of which we are part.

 

Your activity in the protection and transfer of results is also quite remarkable. What are your major milestones in this area?

The group has been granted various patents, various registries of intellectual property and has received various requests from commercial brands. In addition, we have almost daily contact with relevant companies from the technological sector, which often request specific personal knowledge and highly qualified staff to carry out R&D projects.

 

What does CINBIO offer a group like yours in terms of maximising potential and international relevance?

CINBIO gives us the opportunity to collaborate closely with internationally renowned researchers in the field of biomedics. The multidisciplinary work which is carried out in the Centre, improves results and facilitates transfer to the health sector.