Big data and computational modeling have the potential to fundamentally change the way patients are diagnosed and treated. Since the 20th century, diseases and other biological systems have been considered in isolation—which is described as a reductionistic approach.

Today, interdisciplinary researchers are using the wealth of medical data being generated by new technology to take a more holistic approach to healthcare—focusing on biological processes within the patient. Instead of understanding complex diseases by studying their individual components, systems medicine looks at how cells, genes, and tissues interact to cause disease.

Since the term “systems medicine” was first coined in 1992, there’s been an explosion of interest in the field, with the number of papers published increasing by an estimated 41% per year up to 2015. Speakers at the 2nd Conference of the European Association of Systems Medicine 2018, held recently in Utrecht, the Netherlands, presented new computational models for studying complex diseases and assessing patient risk, as well as novel interdisciplinary research.

Coping with Complexity

Understanding complex diseases on a molecular level is among the research topics covered by systems medicine. Complex diseases, such as asthma, Parkinson’s, multiple sclerosis, and cancer, are not caused by a single genetic mutation. Instead, multiple genes, signaling pathways, and regulatory networks are involved.

“That’s why it’s called network or systems medicine,” said Jan Baumbach, Ph.D., chair of experimental bioinformatics at the Technical University of Munich. “We’re not looking at individual gene sets. We’re using big data to look at these systems.”

One challenge is identifying the combinations of genes involved in a complex disease, Dr. Baumbach explained. Even in monogenetic diseases, such as Huntington’s disease, which are caused by a single gene mutation, the disease-causing gene is often expressed the same in patients with and without the disease. In more complex diseases, such as cancer, where multiple genes are mutated, the situation is even more difficult.

Grand Forest
Grand Forest, an ensemble learning platform developed at the University of Southern Denmark and the Technical University of Munich, has been optimized for discovering disease-associated modules from genomic profiling data. The platform may be accessed online, and it provides for supervised and unsupervised analyses of decision trees.

“If you have 22,000 genes, then you can check every gene individually, perhaps check all possible triplets, quadruplets, etc., but the number of possible combinations grows exponentially,” he explained. “We call this ‘combinatorial explosion.’ It’s totally impractical to compute the statistics for all possible combinations, even if your computer was the size of the sun, so you need clever computational techniques to take shortcuts—evaluating the most likely solutions first.”

Dr. Baumbach specializes in using artificial intelligence (AI) approaches, such as machine learning and random forest, to find the most likely combinations of genes. His trick is to pick ensembles of genes that, although not individually linked to a disease, are part of the same molecular pathways. In patients, these groups of genes can work together to drive a disease state.

He also uses unsupervised learning to stratify patients with, for instance, asthma into subgroups from scratch (called de novo endophenotyping) using disease mechanisms, rather than individual genes or gene panels. With this technique, patients are deliberately not grouped by phenotype, but are—instead—clustered in subgroups by an AI.

According to Dr. Baumbach, “If you try a new medicine on 10,000 asthma patients, it might be very effective in 20%, but not effective in the other 80%, so there’s no chance of getting this drug registered, unless you have an effective means of telling the 20% from the 80%. If we can stratify these individuals into subgroups that share common mechanisms, we can check the mechanism against the drug target and look at other treatment options.”

Dr. Baumbach hopes that in the future this technology could be used to tailor treatments to individuals—perhaps by giving multiple drugs at lower doses to reduce side effects or to repurpose drugs that target the molecular networks involved in their disease.

Developing Clinical Tools

Classification of diseases by their molecular mechanisms was also the topic of a talk by Timothy Radstake, Ph.D., a professor of translational immunology at the University Medical Center Utrecht. He discussed using molecular data to develop clinical models of Sjögren’s syndrome and interocular lymphoma.

“We started five years ago to really delve into this niche in a data-driven manner. Most researchers focus on certain molecules, but we decided to use big data to look at multiple subsets of immune cells,” he explained.

Dr. Radstake uses a variety of techniques, including RNA sequencing, proteomics, transcriptomics, metabolomics, and sequencing of the epigenome to classify diseases by their molecular biomarkers. From there, his team has developed a computer model to predict the risk that patients with uveitis will develop intraocular lymphoma.

They developed a tool where they punctured the eye and extracted a little fluid, before doing a simple proteomics analysis. “We think it predicts intraocular lymphoma two years earlier so, if successful, this would potentially lead to earlier treatment for these patients,” he said. The model is now being trialed as a diagnostic aid.

Dr. Radstake is considering moving to clinical trials, but warned that “the real application in clinical decision making is far from reality. We are making the first steps right now, but we have to overcome a lot of challenges.” Among these challenges are carrying out large-scale trials on patient groups from different laboratories. “When we analyze 15 different patient subgroups in the same laboratory, this allows us to have direct comparison, but when we work together with another group, this data isn’t transferrable,” he said. Clinicians specializing in autoimmune diseases are working to overcome these hurdles, but “we’re just at the beginning of how that might work.”

Tackling Big Data

Aridaman Pandit, Ph.D., an assistant professor from the University Medical Center Utrecht, works closely with Dr. Radstake and gave a workshop on big data for clinicians. “We are realizing that medicine can’t turn into personalized medicine without looking at big data,” he said. “But typical clinicians aren’t trained for it like computational biologists.”

The workshop aims to help clinicians understand how systems biology can be used to analyze patient data. Attendees were given simulated immune-cell data and encouraged to perform some of the analyses that Dr. Pandit carries out in his lab. “We want to expose them to a systems perspective, such as how genes interact with each other, so they understand the concept of interconnected networks,” he said. “From there, they can discover the differences between healthy controls and simulated patients.”

Dr. Pandit works on developing computational tools for studying heterogeneity—what makes immune cells differ between patients with, and without, autoimmune disease. “These areas are in their infancy,” he admitted. “Most of things we’re doing are quite novel so, to analyze these types of data, we need to design our own methods.”

He designs novel tools for a variety of projects, including deconvolution and benchmarking approaches to studying immune cells. “Most studies do bulk sequencing of tissue, including skin cells, so if you deconvolute data into different cells, you can see which cell type is most important to focus upon,” he explained.

He’s also applying computational approaches to stem cells, to understand how immune cells form and how the environment affects the types of cells produced. “Our goal is to understand autoimmunity,” he concluded. “One of the main things is the potential [for stem cells] to specialize into different immune cells has been mainly discovered in mouse models or in human models in homeostasis, but in an autoimmune disease there is inflammation inside the patient, and that affects the potential of different cells.”

Understanding the Environment

Another aspect to systems medicine is studying the patient as a whole, including the effect of their environment, or exposome. “There is not much known about how environmental factors affect wellbeing, and in systems medicine people’s focus is mostly on developing personalized medicine and the molecular pathways that contribute to disease,” said Tjitske Bezema, Ph.D., chair of Immunowell and founder of Hidden Health Solutions. She is leading a grassroots initiative to understand how lifestyle and other environment factors affect the health of patients with two autoimmune intestinal diseases—Crohn’s disease and ulcerative colitis.

Working in coordination with researchers from the Amsterdam Medical Centre, Utrecht University, and VU University Medical Centre in Amsterdam, Dr. Bezema is piloting a study of what makes patients enter remission using a technique called Participatory Narrative Inquiry (PNI). “It’s a normal well-used technique in social science, but is not used in this context, so we got together a team of researchers interested in trying it,” she explained.

Hidden Health Solutions data
At Hidden Health Solutions, data about people’s daily experiences are collected and studied using a six-step method called Participatory Narrative Inquiry (PNI). At present, Hidden Health Solutions is using PNI to discern patterns in the experiences of people with chronic illness, and to gain insights about what improves their quality of life. The final PNI step, “return insights,” may occur online or via mailings or meetings.

Recruiting from patient groups on Facebook and by leaving flyers with clinicians, Dr. Bezema had patients write down the story of a time when they felt better. This was accompanied by semi-quantitative and quantitative questions, such as whether their medicine had changed, how much their symptoms improved (on a scale of 1 to 10), and which symptoms got better.

“By collecting experiences from patients in a structured way, and analyzing them, you can see if there are patterns emerging from what people have in common,” she noted. So far, Dr. Bezema and collaborators have collected 65 stories, and from her preliminary analysis she thinks stress and self-esteem might be important topics, as well as exercise and diet. She’s planning to hold a sense-making session in January 2019, with doctors, researchers, and patients, to make a more thorough analysis of the data.

Dr. Bezema also hopes to engage researchers in her project with the aim, in the future, of investigating other co-morbid autoimmune disease and to spark interest in further researchers. “If stress is an important factor,” she suggested, “we can find out what sort of stress may call for what sort of stress management techniques. Then, perhaps, we can see if we can develop interventions that work alongside medicine.”

This site uses Akismet to reduce spam. Learn how your comment data is processed.