Our new guest is Miha Moškon, Associate Professor at the Faculty of Computer and Information Science, University of Ljubljana, and a member of the Computational Biology Group at the same faculty. His passion is studying biology from the perspective of a computer scientist: “One of the most rewarding things in my work is to study a biological mechanism in detail, replicate its dynamics in a computational model and then observe it in a computer simulation. In other words, to replicate living things inside a computer”.
That is indeed fascinating. “I was always interested in different fields but then decided to study computer science. I did my bachelor thesis on a computer model of collective behaviour of birds. Later on, my work shifted to a smaller scale of biological systems, which was initiated with also my participation in synthetic biology iGEM competitions, where the Slovenian student team has been very successful. I did my PhD in the field of computational modelling of engineered biological systems as an alternative to digital electronic circuits. We are still working on the computational design of biological systems with information processing capabilities in the context of synthetic biology, but recently our work has shifted more towards systems biology, where similar approaches are not used to design novel but to study natural biological systems”.
Miha Moškon is a full-time teacher at the university, even though he mostly regards himself as a researcher: “We are currently working on computational approaches for automated integration of large-scale omics data into different types of computational models. Examples of these include genome-scale metabolic models and Boolean models of gene regulatory networks.
Data-integration approaches allow us to adapt computational models to a specific biological context by using a reference model together with big data, usually omics data. For example, we can use a dataset gathered from a specific group of individuals to reconstruct a disease-specific model. Or we can use a dataset from a specific individual to reconstruct their personalized model. The models produced in such way are not only useful for the interpretation of given datasets, but also for the generation of new data, new hypotheses, their testing, etc. For example, we can build a computational model of a specific disease and then test what happens if a certain perturbation occurs within the network. A perturbation might be represented by a change in a diet, silencing a gene or a drug intervention”.
The approaches Prof. Moškon is working on can be applied to a wider scope of systems biology, but are also used to answer questions in medicine, especially in the context of complex-systems diseases. “As the name suggests, these are systems disorders, which means that their occurrence cannot be pinpointed to a single cause, but are a combination of several factors, which are hard to identify, unless addressed from a system perspective. This represents a shift from a reductionist approach (each part of the system is analysed separately) to an integrative approach (a system is more than a sum of its basic parts). Computers have allowed us to study biological systems from such a perspective. The field has made significant progress in recent years especially due to the development of different high-throughput experimental techniques, which are able to generate large amounts of data, and with the development of computational power and algorithms, which can be used to analyse these data very efficiently”.
Many scientists agree that Modeling and Simulation could be a game changer for the healthcare industry, but in what ways? “Currently we can reconstruct models of different cellular processes, but we are still coping with connecting these models to the whole systems. For example, to obtain a whole-cell, whole-tissue or even whole-body (whole-organism) models that would be scalable from simpler to complex organisms. This is essential for the future progress of the field since no biological process works in isolation and the predictive power of ‘isolated’ models is limited. For example, to accurately describe the metabolism, we must also be able to connect its model with a model of gene regulatory and signalling networks. The modelling approaches used in each of these networks are advancing every day and so are the approaches for their integration into more comprehensive models. I am confident that we will be able to address the model integration problem soon”.
“In the context of medicine, this will allow us to reconstruct a digital twin, which will represent a very accurate virtual representation of the human body. Of course, every human organism is different. In this context a reference digital twin model will have a potential to serve as a reference or a ‘scaffold’ model. Having appropriate algorithms and enough computing power will allow the physicians to adapt a scaffold model to obtain a patient-specific model in real-time at a point-of-care using patient specific-data. Practical applications of this can be a game changer for the healthcare industry. A physician will acquire the patient’s data (e.g., omics data) and will integrate them into a scaffold model to obtain a personalized model. The latter will then be used to identify the anomalies in the network (diagnostics), to identify the interventions that would bring the system back to its healthy state (personalization of therapy) as well as to simulate what would happen if certain interventions were or were not issued. Moreover, digital twins will be used to accurately track the progress of a selected therapy and to adapt this when necessary”.
One of the main challenges that the healthcare industry is facing today is the availability of data in standardized machine-readable formats: these are very sensitive data, so this issue needs to be addressed carefully and in long term. Another problem is the readiness of the healthcare community to adopt computational modelling approaches: “The practitioners of medicine are in my opinion conservative towards the use of these approaches, especially due to the knowledge gap between the medical education and the computational and mathematical sciences. We can address this gap by introducing different computational and bioinformatic courses to the curricula of medical faculties. This would allow us to educate a new generation of medical doctors that will be able to address complex nature of complex system diseases more efficiently. And maybe we will soon have a new medical specialty, such as a computational physician”.