Saez-Rodriguez Group

Systems Biomedicine

Comorbidities in Heart Failure – A Network Approach

Chronic heart failure (HF) is the permanent inability of the heart to pump blood sufficiently through the body to meet its demands for oxygen and exchange of metabolites. The three most frequent symptoms that this condition provokes are shortness of breath, fatigue and fluid retention, which can cause a significant loss in the patient’s quality of life. Furthermore, progressive HF has a poor outcome leading to a 1-year mortality rate of roughly 50% in patients with severe heart failure, who exhibit symptoms even while at rest. Importantly, heart failure often presents with comorbidities, including diabetes, renal dysfunction, and depression. Despite decades of research, a comprehensive understanding of HF, including the role that comorbidities play in development, progression, and outcomes in HF remains incompletely understood.

We are evaluating the role of multiple comorbidities in the clinical manifestation, treatment, and outcome of HF patients. Our goal is to improve understanding of HF with a systems medicine approach (Figure 1).  This project is a close collaboration between the Saez group and the Jobst-Hendrik Schultz group at the Center for Psychosocial Medicine (head of department: Prof. Dr. med. Hans-Christoph Friederich), Heidelberg University Hospital. This project is supported by the Klaus Tschira foundation through the Informatics for Life consortium. This consortium was formed in 2018 to embark on a multidisciplinary approach to study cardiovascular diseases by joining forces of clinical and computational experts.

Figure 1. Schematic of project outline.  Molecular signature extraction and contextualization will be followed by building a comorbidity network incorporating clinical data. Together these elements will be used for association and prediction of clinically relevant outcomes.

 

Deriving a Molecular Signature for HF

Mechanistic signatures from omics data of HF and its comorbidities will be extracted using different approaches that use prior knowledge in the form of biological networks. These methods, such as PROGENy (Schubert et al. 2018), DoRothEA (Garcia-Alonso et al. 2018), Omnipath (Türei, Korcsmáros, and Saez-Rodriguez 2016) and CARNIVAL (Liu et al. 2019), contextualize molecular profiles in functional elements such as pathways and transcription factors. These elements can be associated with other layers of biological complexity or clinical information. Moreover, statistical models will be applied to these signatures to generate a connectome of recurrent molecular HF events with other diseases or drug responses.

 

Identifying Clinical Comorbidity Networks

Clinical data resources provided by the Heidelberg University Hospital will be used to build a disease network around HF to clarify the relations between comorbidities like chronic kidney disease, hypertension, inflammatory diseases and diabetes. We will also evaluate psychiatric disorders, as some of these have been closely linked to HF outcome. Clinical data will be used for association analyses and longitudinal studies, and will also be used as the basis for development of comorbidity networks.

 

Predictive Modelling of Heart Failure

Using insights gained from our multi-omics and clinical network approaches, we will perform machine learning and predictive modeling, with the goal of identifying ways to improve the diagnostics and treatment of HF with multiple comorbidity.

In addition, to efficiently develop computational strategies to predict HF, a crowdsourcing project will be set up in the form of a DREAM challenge. Through this challenge, any scientist in the world can develop methods to predict HF from the clinical data, by submitting algorithms to be submitted benchmarked against the data without directly accessing it. The outcome of this process will be the development of many algorithms, that will be then integrated to generate an ensemble.

 

Selected References:

Garcia-Alonso, Luz, Francesco Iorio, Angela Matchan, Nuno Fonseca, Patricia Jaaks, Gareth Peat, Miguel Pignatelli, et al. 2018. “Transcription Factor Activities Enhance Markers of Drug Sensitivity in Cancer.Cancer Research 78 (3): 769–80. https://doi.org/10.1158/0008-5472.CAN-17-1679.

Liu, Anika, Panuwat Trairatphisan, Enio Gjerga, Athanasios Didangelos, Jonathan Barratt, and Julio Saez-Rodriguez. 2019. “From Expression Footprints to Causal Pathways: Contextualizing Large Signaling Networks with CARNIVAL.” BioRxiv. https://doi.org/10.1101/541888.

Schubert, Michael, Bertram Klinger, Martina Klünemann, Anja Sieber, Florian Uhlitz, Sascha Sauer, Mathew J. Garnett, Nils Blüthgen, and Julio Saez-Rodriguez. 2018. “Perturbation-Response Genes Reveal Signaling Footprints in Cancer Gene Expression.” Nature Communications 9 (1): 20. https://doi.org/10.1038/s41467-017-02391-6.

Türei, Dénes, Tamás Korcsmáros, and Julio Saez-Rodriguez. 2016. “OmniPath: Guidelines and Gateway for Literature-Curated Signaling Pathway Resources.” Nature Methods 13 (12): 966–67. https://doi.org/10.1038/nmeth.4077.