Newsletter #1 - Meet the iASiS Team

Interview with Maria-Esther Vidal – a technical perspective

Maria-Esther Vidal

Maria-Esther Vidal is a visiting Professor at the Leibniz University of Hannover (LUH), and Head of the Scientific Data Management group at the Leibniz Information Centre for Science and Technology and University Library (TIB) in Germany.

What is your role in iASiS?

Our role at LUH is to develop all the necessary techniques for integrating heterogenous data sources into a knowledge graph. We are also semantically describing the data in these data sources and exploiting the semantic description in order to integrate data that looks different, but in the end corresponds to the same entities in the real world.

Tell us a bit about your background and the work you do at LUH.

My background is in databases and semantic web, and particularly the integration of data. We have developed a technique to integrate data and use the semantics of the data during this integration. Our work also includes query processing and knowledge discovery, because we have to integrate all this data into a knowledge graph, and then several applications have to be developed in order to extract knowledge from this knowledge graph

What is your motivation for this project?

I have worked in the biomedical area for several years, and, so far, the projects I have been involved with apply these data integration techniques to join up different data sources but don’t try to transform the data into knowledge. This idea of transforming data into knowledge and working on two very deadly diseases like lung cancer and Alzheimer’s was very motivating for me. I’ve lost people very close to me to these two diseases, and I would like to use all my knowledge to contribute towards understanding these diseases.

What do you think will be some of the biggest challenges for this project?

Integration of data, extraction of knowledge, and representation of that knowledge. When I say integrating data I don’t mean just from the technical point of view of pulling together the datasets, but resolving interoperability issues such as inconsistencies. One of the main problems that has to be addressed when you process big data is the problem of the quality of the data. Inconsistencies and ambiguities will always be present in the data, and we and the tools we develop have to deal with them.

What kind of data are you working with? How do you think this data can help patients?

We are working with different types of datasets. We have open data, for example databases that describe somatic mutations and information on the genes where these mutations occur, the drugs to which these mutations are resistant, and the tissues where these mutations have been observed. We also have open datasets describing drugs and the interactions between drugs, the side effects, conditions for which these drugs have been prescribed, as well as clinical trials results. We also work with control data, where we have clinical records. This data is anonymised before we receive it, so we don’t have the raw data that identifies patients, but we have an entity that represents a patient, so for that entity we have demographic data, the treatments they have undergone, and the different stages in this person’s clinical record. And one thing that is extremely important to identify from this data is, for example, the long survivors with lung cancer or Alzheimer’s disease. There isn’t a clear understanding of why there are some patients that, with the same characteristics as others, live longer. So, we are applying analytics tools such as community detection algorithms, in order to identify the communities of people with similar characteristics and identify these long survivors.

What do you consider as critical for the success of LUH’s part of the project?

Collecting enough data. If we don’t have the necessary data, we won’t be able to accomplish our objectives. And I need to have a clear understanding of the meaning of the data from the different partners. This will help tremendously with reconciliating the ambiguities and inconsistencies we see in the data. It is very important that we work closely with the clinical partners to explain these issues, so in the end we have the best possible dataset to work from.

What do you hope for as the end outcome of iASiS?

That in the end I can help clinicians understand patterns and improve patient survival. This is my dream. My dream is that at the end of the project we can characterise the long survivors in lung cancer and Alzheimer’s disease. That we can maybe discover unknown relations between the characteristics of the patients and for example previously unknown mutations, which could help us choose the right treatments for the right patients and inform health policies.

Interview with Maria Torrente – a clinical perspective

Maria Torrente

Maria Torrente is Head of Clinical Research at Puerta de Hierro University Hospital in Madrid.

What is your role in iASiS?

I am a clinician and researcher at the Medical Oncology Department at Puerta de Hierro University Hospital in Madrid. We are the data providers for the consortium regarding lung cancer patients.

Tell us a bit about your background and the work you do at Puerta de Hierro University Hospital.

As a doctor, l have worked in different hospitals in different cities (New York, London and Madrid), where l received training not only as a clinician but also as a researcher. I have been involved in many clinical research projects and clinical trials, both national and international. Here at the hospital my work has always involved both clinical and research work, that means that not only have l worked with patients but also as a researcher in clinical projects (protocol elaboration, writing of proposals, active member of different hospital research networks and scientific societies), focused in the patient’s healthcare as well as in ways of obtaining funding in order to develop new projects that will help improve clinical practice, patient’s care and outcome.

What is your motivation for this project? What do you think is the most exciting aspect of the project?

My main motivation is to be able to address a fatal disease such as lung cancer from a completely different and new approach, which is big data analytics. It is also an \enriching experience to be able to work and collaborate with different partners within an European consortium, most of them technical partners, which makes the project even more challenging

What do you think will be some of the biggest challenges for this project?

I think there are 2 main challenges: 1. Collaboration between clinicians and technical partners, and 2. Obtaining results that are relevant for the disease and help to improve clinical practice, patient’s survival, treatments and reduce toxicities.

What area are you focused on as part of iASiS? What scenarios are you testing?

Lung cancer. As there are so many unanswered questions in oncology, many regarding lung cancer, we have focused in finding differences among patients with similar characteristics. Lung cancer is a very heterogenous disease, which means that although many patients have the same disease characteristics (regarding type of tumour, stage, treatments) they usually respond differently to treatments, develop different kinds and grade of toxicities, and therefore have different outcomes, especially in terms of survival. Therefore, we aim to analyse those patients with better response to treatments, less toxicities and greater survival, and compare them with the rest of the patients, in order to seek for patterns that may guide us to improve clinical practice and improve survival in most of the patients with a more personalised treatment and follow up.

How do you envisage iASiS will help clinicians like you in the future?

To be able to analyse huge amounts of data regarding all clinical variables, not only the obvious and usual ones, as they usually don’t answer most of our doubts during a patient’s treatment.

Why is this approach needed for lung cancer? Why were you keen to be part of this?

Lung cancer treatment has improved greatly in the last 10 years but there is still room for improvement regarding treatments, outcome, toxicities, which remain unknown or are quite new areas. Also, it is important to access daily practice data, which is very different from that obtained in clinical trials, which only regards certain clusters of patients. Large registries are urgently needed as well as risk scores for patient stratification and algorithms for better management of patients.

What could this mean for patients?

Better management of their disease, patients empowerment, and better survival.



In the spotlight

Meet the iASiS Team

Objectives and progress in our first year


Future Events