“The good physician treats the disease; the great physician treats the patient who has the disease” (Sir William Osler, 1903)

Personalization & Precision Medicine

Current medical practice relies on a catalogue of diseases, each defined by pathophysiology, symptoms and outcomes. In case of strict causality a specific diagnosis has one clinical consequence and this “action-reaction” scheme then applies for all subjects affected. Inter-individual variance in disease progression and response to treatment is fairly limited. For many, especially chronic and age-associated diseases however the situation is more intricate. While symptoms, treatment and outcome still hold true for a cohort, we observe individual heterogeneity in disease progression as well as response to therapy. In this scenario precision in diagnosis and treatment needs improvement by fostering stratification and personalization, particularly if different treatment options are available.

The latter is the aim of the collaborative R&D initiative DC-ren in an extremely relevant disorder: Diabetic Kidney Disease (DKD). DC-ren, the abbreviation for “Drug combinations for rewriting trajectories of renal pathologies in type II diabetes”, has started in 2020 with a runtime of 5 years. DC-ren has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 848011.

Diabetic Kidney Disease

DC-ren focuses on DKD, a severe long–term complication of type II diabetes mellitus (T2DM), which is characterized by alterations in glucose metabolism caused by insulin resistance. DKD, a gradual loss of renal function leading to end stage renal failure, is per se a complex disorder, which is further modulated by the high burden of co-morbidities. Complexity is also evident on the level of pathophysiology and guidelines recommend drug combination therapy. In recent years treatment choices have increased, with several novel drugs with proven renal as well as cardiovascular benefit entering the market. Patients with DKD display remarkable individual variability in disease progression and response to certain drug combinations. Thus, there is an evident clinical need for adding precision in treatment via personalization. The central objective of DC-ren is to provide a decision support technology for selecting the optimal drug combination treatment to improve the prognosis on the level of individual patients with DKD.

While pursuing a clear focus on DKD we are confident that our novel methodological approach has the potential for serving as blueprint for adding precision in treatment of complex disorders.


Disease & treatment: a dynamical system

Guidelines recommend regular monitoring of renal excretory function in T2DM by a laboratory parameter called “eGFR”. Of note, not all patients with T2DM develop DKD, and those who do loose eGFR with a remarkable inter-, but also intra-individual variability over many years. This is due to the complex interaction of genetic predisposition and environmental factors that results in a personalized trajectory of development and progression of DKD.

Formally DKD evolves as a sequence of pathophysiological states being causally linked by action-reaction. Certain trajectories resemble a fast decline of eGFR while others result in a fairly stable situation. Inter-individual heterogeneity is further complicated by intra-individual variability. Some states may result in stable or even improving renal function, others in rapid decline and some may see rapid transition while others are stable for years. A state is defined by its underlying pathophysiology, and we have to keep in mind that this is the point of action of drugs. Therefore drugs exhibit beneficial effects only in certain states, while they fail in others. When applying such a state model we can capture inter-, but also intra-individual variance in drug response.

DC-ren follows the concept of individualized trajectories of state transitions with state-specific drug response according to underlying pathophysiology. Such a setting is conceptually well established in the scientific domain of dynamical systems research.


According to a dynamical systems approach (individualized trajectories of state evolution combined with state-specific drug response) the DC-ren work plan takes care of three central R&D aspects.

  1. Detailed longitudinal follow-up of individual patients: Conventional approaches retrieve clinical and molecular characteristic for a cohort of patients at baseline, and clinical outcome at one follow-up time point to assess drug response. In contrast, DC-ren has access to baseline and several follow-up time points for a significant number of patients. This allows screening for states and state transitions, and realization of state-specific prognosis and drug response assessment.
  2. Novel analysis strategy: Reference implementations follow a top-down approach by starting with a data space capturing a cohort of patients. Analytics aims at statements to be informative for the cohort, subsequently being applied to individuals. DC-ren switches the strategy to bottom-up stratification by gradually building consensus trajectories for subgroups of patients for which statements on prognosis and drug response are retrieved.
  3. Foundation: Predicting the consequence of a state-specific drug effect on disease progression demands insight into mechanism (action-reaction). DC-ren follows a novel integration of explorative analysis (statistics, machine learning) and rule-based simulation to allow predicting optimal drug combinations on a state level for the subgroups of patients being assigned to such states.


Application goals

The translational focus of DC-ren is to establish and validate a decision support software application for optimizing combination drug therapy in patients with DKD on a personalized level in a “technology readiness level 6” prototype (i.e. allowing practical demonstration in a relevant environment).

For implementation a validation study, leveraging on existing clinical trial repositories and realized as a “virtual trial”, will test if technology-mediated decisions add in precision in personalized drug response prediction when compared to present clinical guidelines. This validation is pivotal for moving the decision support solution in higher technology readiness levels, and ultimately in clinical settings.

    On top, we test the decision support solution in two relevant application scenarios, namely in DKD drug R&D and compound recovery.
  1. Clinical testing of novel compounds or repositioning of available medication needs optimal definition of the target patient cohort for assuring the best possible risk/benefit ratio. In analogy to approved medication this demands rational evaluation of the patient-specific disease pathophysiology and the compound mechanism of action. We will test if our decision support software application allows patient stratification for these settings.
  2. Recovery of active compounds, which were tested in DKD but failed in efficacy and/or safety. In this situation alternative stratification may identify patients still benefitting from such drugs and we will test in a selected example if the decision support application enables such recovery strategy.

Science goals

All science in DC-ren is geared towards implementing the technology solution. The toolbox elements come from analysis of a high-dimensional data space covering clinical and molecular characteristics for a substantial cohort of DKD patients. Here we need to capture details of the personalized disease pathophysiology at the interface of mechanism of action of drug combinations on a common layer of molecular processes, which interact as a System-of-Systems. The challenge is to identify molecular biomarkers and clinical parameters being informative on drug effect in dependence of underlying process configuration for state-specific pathophysiologies. For coming up with an ideal biomarker panel we combine in-silico based candidate biomarker selection with an explorative proteomics approach. The data space shall serve for generalizing and inferring the consequence of the drug effect on personalized (state-specific) pathophysiology, ultimately allowing statements on optimal drug combination matches.

Statements on the level of the toolbox are derived from a hybrid AI. “Hybrid” refers to a combination of data-driven generalization from statistics, machine learning and rule-based inference from simulation. Such an approach shall allow us to streamline probabilistic predictions with causal (action-reaction) statements on drug response. “Data rich” analysis up to learning has become highly acclaimed in recent years also for solving medical challenges. We believe this methods compendium needs expansion with rule-based, mechanistic approaches, as causality and determinism are pivotal for bringing decisions on optimal drug combinations forward in precision medicine.


For the public

04/2020: ”Computer model aimed at optimising treatment of diabetic kidney diseases.” (project start press release as pdf)

03/2020: European Union CORDIS fact sheet (link)

For professionals

04/2020: ”Wrong biomarkers or wrong questions?” (article as pdf)


Interdisciplinary R&D

Contemporary clinical research critically depends on interdisciplinary teams, and DC-ren balances theoretical, experimental and clinical expertise. The DC-ren analytical concept routes in applied Category Theory, a framework providing concepts and tools for modeling dynamical systems. Its specific strength is the representation of composite function, in DC-ren being distinct DKD pathologies at interference with drug mechanism of action. Translating the concept via combining statistics and simulation on the basis of a rich clinical and molecular data space allows taking a novel perspective for assessing personalized drug response as hybrid AI solution. We leverage on extensive biobanks and clinical data repositories readily capturing personalized response to various drugs approved for DKD. We complement clinical phenotyping with top-level molecular profiling, going for multiplexed assays and proteomics. This will provide us with a targeted data matrix for capturing disease mechanisms and drug impact at the effector level of molecular processes: A System-of-Systems. All analytical and experimental work is embedded in strong clinical research for assuring technology development right in focus of patient needs. Practical implementation follows agile software development tailored at reaching a prototype decision support platform, finally evaluated in dedicated validation and DKD drug application studies. With this setting DC-ren bridges from a novel concept to a clinical solution with substantial technology readiness level.


Get involved

We are committed to collaborative R&D, and cordially invite stakeholders to open discussions eventually leading to joint initiatives.

  • While Dc-ren pursues a translational Proof-of-Concept work plan, all our activities are geared toward improving treatment of DKD patients. Clinicians of DC-ren see patients on a daily basis, and we encourage patients and patient organizations to link with us.
  • Personalization, be it novel methods for patient stratification or adaptive design of clinical studies imposes challenges on regulatory procedures. We actively link with regulatory agencies together with experts in health technology assessment for moving personalization strategies to a next level of implementation.
  • You are involved in developing novel compounds for treating DKD? Get in contact for hearing more about our patient stratification approach – and apparent collaboration opportunities for your drug R&D.
  • You are in basic science disciplines as listed above and want to explore translation opportunities? Link with us! The DC-ren team covers a wide range of academic disciplines, including applied computer science, statistics & simulation, omics & biomarkers, all embedded in clinical research and complemented by software architecture design and implementation.

DC-ren contact

The collaborative research project DC-ren is coordinated by Univ. Prof. Dr. Gert Mayer, Medical University of Innsbruck, Austria.

Please use the following means for interacting with DC-ren:


Gert Mayer
Department of Internal Medicine IV (Nephrology and Hypertension)
Medical University Innsbruck
Anichstrasse 35
6020 Innsbruck

DC-ren funding:
DC-ren has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 848011.