Science and Innovation

Interview by Sergi Macià i Muñoz for Barcelona Science and Universities and Núvol.

Clara Prats Soler, born in Barcelona, is the daughter of a biologist and a musician, professor at the Polytechnic University of Catalonia, coordinator of the Computational Biology and Complex Systems group and member of the Institut d’Estudis Catalans and the Royal Academy of Medicine of Catalonia. During the coronavirus pandemic, she has been a well-known media figure who kept the country informed of the situation.

Clara, in addition to teaching and research, we have seen you regularly in the media. How have you experienced this media boom?

In fact, it all started on the radio, on RAC1 with Xavi Bundó. The UPC group was asked to attend regularly on Saturday mornings to explain what was going on in the world. Everyone was a bit lost… I was trying to translate a complex situation to society, who didn’t need to know or have a basis in epidemiological dynamics or the concepts behind it.

You have become a media scientist…

In fact, for me, media exposure is not a subject that has ever caught my attention, but at that time, as a research group, we thought it was important. If we wanted people to be aware of epidemiological behaviours, it was important to carry out this task of communication, of dissemination, of explaining concepts, of explaining above all the reasons for the restrictions. Numbers, if they are not understood, do not help society.

Before the pandemic, with your research group, you were already making mathematical models to explain complex biological behaviour. Since when have these models been used?

I could talk for an hour now (laughs). Mathematical models have been around for a long time. However, since the 20th century, with the new technologies, many doors have opened up for us: as computers have become more powerful and databases have grown in scope, we have been able to solve more complex problems. I always tell my students that physics is very simple – and then they look at me with a look of disbelief – but biology is much more complex! Things that in physics can be answered with a paper, a pencil and a calculator, in biology you need a computer with a lot of power and a lot of data to analyse.

But how reliable are the resulting predictions?

Prediction is the last step in a chain. The model, from the outset, serves to better understand a given system. You make some hypotheses, you formulate these hypotheses in equations, in algorithms, and when they describe the dynamics of a system well enough – we have parameterised it, we call it – we put it through a validation phase, where we check the predictions against the data. In this way, we can understand the degree of uncertainty in the model’s responses, since any prediction is, by definition, imprecise.

How accurate have the predictions been in the pandemic?

During the first wave, we could predict two to three days ahead. As we have learned more about the epidemiology of SARS-CoV-2 and the data have become more robust, the window of reliable prediction has increased to two, three or even four weeks, depending on the situation. However, technologically there are still things that are very unpredictable, such as when a wave will start to grow. Therefore, at this point, the prediction is again a few days.

Artificial intelligence (AI) is in the spotlight these days. To what extent can AI be applied to predictions?

Artificial intelligence tools are very powerful, but I think they should be used in coordination with mechanistic models, because in the end these tools are black boxes that allow us to make a prediction, but they don’t explain why. When you build a model, you build it based on hypotheses, generally on a much finer knowledge of the system. I would look at AI as a complementary tool.

In fact, we have seen that Chat GPT prefers to invent an answer rather than admit that it has none…

Exactly, and hence one of the modeller’s jobs is to discern whether the information obtained comes from a reliable biological basis.

Changing the subject, you have sometimes commented that during the pandemic you used indicators that were easily understood by society, such as the number of people infected, the number of sick people, the number of hospitalisations and the number of deaths. Is this information used to organise the health system better and avoid overcrowding?

It is one thing if we can get the information and another if we can avoid saturation. Predictive models are constantly being used for influenza or bronchiolitis in paediatrics, for example, which can help to design immunisation campaigns, manage bed availability or reinforce health staff in primary care centres. But these predictions are not used today to articulate measures of population behaviour, as was done during the pandemic.

Since the pandemic, has the profile of the hospital engineer or epidemiologist been consolidated?

The entry of new non-health professionals into the healthcare world is a complex issue that is on the table right now. Not only for providing models, but also for data analysis or diagnosis with artificial intelligence tools. The pandemic has allowed us to create collaborative networks that have benefited from the integration of these professionals… This works very well. There is a way to go, and we have been able to see that it is a necessary change. At GIPS, the Interdisciplinary Group of Professionals Linked to Health, an initiative of the Barcelona College of Doctors, we are working precisely on everything related to the incorporation of new professional profiles, and I am the president of the Management Committee of this group.

Collaboration networks?

Yes, for example, in paediatrics, we participate in a network of paediatricians that brings together more than 150 or 200 professionals from all over Catalonia: COPEDI-CAT. It was created during the pandemic, but we have maintained it and we share concerns and doubts about the evolution of emerging diseases and we carry out interdisciplinary research projects. We work every day to try to answer these questions through data analysis and modelling.

For the last two years, the Department of Health has been promoting, through SIVIC, the new Information System for Infection Surveillance in Catalonia, which was created as a result of the work done during the pandemic, the parallel creation of the Epidemiological Intelligence Network of Catalonia, XIEC. It is not yet consolidated, but work is being done to create this network of groups that work on modelling and that can answer relevant questions about what has happened or what may happen. We are meeting weekly or fortnightly with public health personnel from Catalonia.

What do other countries do, do you share information?

I don’t know much about how each country organises itself internally. What I can say is that from the ECDC (European Center for Disease Prevention and Control), the European agency for the control of infectious diseases, from the beginning of the pandemic they tried to promote the whole issue of the collaborative network between modellers. We have been participating in this European network for almost two years. Each institution contributed predictions at the local level and, on a weekly basis, with all these data, a joint prediction was made: a prediction based on the predictions. This is working very well. In fact, it is now being extended to the prediction of other seasonal respiratory infections.

We have talked about the entry of non-healthcare workers into the healthcare world, and we have seen how the work you do is at the confluence of engineering, biology and medicine. Today, how can a young person prepare for that kind of work, which seems to be on the rise? What training do they need to have?

You can join it from almost any branch of science. In health, there are biomedical engineers, doctors, human biology graduates, telecommunications engineers… We need interdisciplinary groups. I am in favour of interdisciplinarity in science: it enriches us enormously.

And the million-dollar question: Clara, are we prepared for a new pandemic?

I hope that the next pandemic will be decades away… (laughs) Pandemics occur on average once every century. We need to see if these collaborative structures that are being created now will last over time so that we can use them in the next pandemic. Today, I think we are better prepared than we will be in 2020.

And finally, I know that apart from being a physicist, researcher, creator of predictive mathematical models and television broadcaster, on a more personal level, you are the mother of two children and you have directed a choir. How do you combine this?

I think there’s one overriding factor, which is to assume that you’ll be less productive at work and to assume that you won’t be the mother you hoped you would be. In other words, there is no such thing as superwoman, the concept of a woman or man who can do everything. The day has 24 hours: you have to make compromises and let yourself be helped.

It is said that men are often listened to more than women on equal terms. Have you ever heard yourself less listened to because you are a woman?

I don’t have the perception that this has conditioned me. But it is also true that I don’t know if I would perceive it if it happened to me… My mother did her doctoral thesis when I was a teenager and, therefore, I have never thought that I couldn’t do it too. But I also don’t know to what extent I have stopped giving my opinion at certain times, perhaps due to a lack of ambition or what they call nowadays the imposter syndrome…

And beyond that, do you think that the fact of being a woman has been detrimental to your professional career?

The first moment when I realised that being a woman was detrimental to me in certain aspects, although I wouldn’t change it, was when I had children. It’s a biological issue: there are pregnancies, which are complicated at times. When you don’t sleep because you are breastfeeding at night and your head doesn’t work the same way… Where before you thought for one, now you think for three… There are many things, but in terms of the day-to-day work, looking after the little ones and accompanying them, it is necessary to continue working so that they are shared with the couple.

Finally, what advice would you give to young people who are considering a career in science and technology and becoming future researchers?

For me, the most important thing, and this applies to young people who want to work in research or in any other field, is the team of people. You have to feel good, they must treat you well, they must look after you, they must accompany you… For me, teamwork is the basis of everything. I don’t think it’s not so important where you end up working, but who you end up working with.