If it were possible to understand and diagnose why some patients do or do not respond to a given treatment this could significantly reduce suffering, and potentially also costs of health care and drug development. This is complicated by the involvement of thousands of genes and gene products in common diseases like asthma, obesity and diabetes. Another problem is that treatment is generally started late in disease processes, when there are sufficient symptoms and signs to reach a diagnosis with conventional methods.
High throughput technologies allow the analysis of all disease-associated genes and gene products in samples from patients and controls. However analysis of the resulting data is major challenge. Systems medicine is an emerging discipline that aims to address this challenge by combining high throughput, computational, functional and clinical studies. Briefly, disease associated genes identified by high throughput technologies are computationally mapped on network maps of human protein interactions. The most important of those genes tend to co-localise and form so-called modules.
The modules can be further analysed to find pathways and key disease genes. The Center for Personalised Medicine, CPMed, has recently described a strategy to define such modules and how to identify diagnostic and therapeutic candidate genes, which were validated by functional and clinical studies (Bruhn et al. Science Translational Medicine 2014). Article: A Generally Applicable Translational Strategy Identifies S100A4 as a Candidate Gene in Allergy.
We have also shown that modules have the potential to individualise medication (Gustafsson et al. Genome Medicine 2014), as well as find markers for early diagnosis, even before symptoms occur (Gustafsson et al. Science Translational Medicine 2015). Article: A validated gene regulatory network and GWAS identifies early regulators of T cell–associated diseases
Importantly, we found that one module was shared by many different diseases. This shared module is likely to have an important disease-causing role. It is enriched for disease associated pathways, polymorphisms identified by genomewide association studies, as well as diagnostic markers and therapeutic targets (Barrenäs et al. Genome Biology 2012, Gustafsson et al. Genome Medicine 2014). We hypothesise that genes in the shared module, as well as in specific modules may contain important diagnostic markers for personalised medication.
We are therefore studying multiple diseases in order to characterise both shared and specific modules. We are doing this in T-cells from patients with T-cell associated diseases. Functional validation studies are performed in T cells from healthy controls, as well as in mouse disease models. The principles are described in a short introductory article: Targeted omics and systems medicine: personalising care (Zhang et al. Lancet Resp Med 2014) and a more comprehensive review article: Modules, networks and systems medicine for understanding disease and aiding diagnosis (Gustafsson et al. Genome Med 2014).
We currently study the following diseases
- multiple sclerosis
- seasonal allergic rhinitis
CPMed is a collaborative effort that includes clinical specialists and a core systems medical team with integrated expertise in genomics, epigenomics, bio-informatics and clinical translation. This is facilitated by participation in the leadership of an EU project, CASyM, which aims to implement systems medicine in European clinical research and practice. Mikael Benson is advisor to the European Commission Health Directorate for Horizon 2020, as well as PerMed an EU initiative for individualized medicine.