Atrial fibrillation (AF) is a major cause of stroke, heart failure and long-term disability in the elderly in Europe. However, the efficacy of current therapies is limited, leaving a large portion of the AF-related morbidity and mortality untreated. In addition, there is a huge disconnect between our current understanding of mechanisms underlying AF and the clinical management of patients. It is now admitted that AF is the clinical expression of an atrial cardiomyopathy. This atrial cardiomyopathy is characterized by a long period of clinically silent progression, with a number of environmental, individual or local factors acting together on disease progression and, most often, a dramatic acute clinical expression. Hence, there is an urgent need for personalized and up-stream therapy to prevent AF burden and stroke. In order to tackle the challenge, new tools must be developed to both identify the various pathogenic processes contributing to the multiple faces atrial cardiomyopathy and be accessible to the general population in Europe. The over goal of the proposed network will be to bring together recent developments on the clinical imaging of the atria and their relevant algorithms with machine learning approaches in order to develop new diagnostic tool of the atrial cardiomyopathy. Large data collections will be constituted by (i) results of the transcriptomic analysis of a large collection of samples of human atrial and performed within the H2020 CATCH ME network and that will be used to develop new mechanism and etiology specific imaging of the atria and (ii) of a large number of atrial MRI imaging collected in the various centers of the network and that will be used to validate the new imaging parameters and then for the machine learning approach. The network will bring together European leaders in the field of atrial cardiomyopathy and AF, on cardiac imaging (Oxford and Birmingham Universities, Paris IHU ICAN, Madrid CINIC) and AI (Sorbonne University). The European network will be the right scale to validate new algorithms and AI in large cohort of patients from the various centers. The network will benefit from the expertise of partners in large clinical study on AF and patients cohorts (A. Goette, P. Kirshoff). The network will be under the umbrella of the ESC (B. Casadei will be a partner). The CMDO, Canadian research network in Cardiometabolic health, diabetes and obesity from Sherbrooke University will also join the network. Justification: The proposed project follows previous FP7 and H2020 networks on AF which one major achievement is the concept of “atrial cardiomyopathy”. The proposed network will pursue the sustained effort of the European Consortium to tackle this major health problem.

The generic call for proposals from the ANR 2015 clearly encourages and supports integration of reliable, multi-sourced data that, when efficiently analyzed, interpreted, disseminated and used, can become an almost inexhaustible source of knowledge to fuel a learning healthcare system. This is exactly the purpose of ProteoCardis which proposes an unprecedented intestinal metaproteome-wide association study of coronary artery disease (CAD), where newly acquired huge metaproteomics data (several thousands of bacterial proteins and several tens of thousands of peptides) will be put into perspective with patient records, metabolic features, complete cardio-vascular exams and outcomes that are acquired in the FP7 MetaCardis framework and will be refined with the ANR’s support. In the MetaCardis FP7 framework (2013-2017), the gut metagenomes of more than 2000 patients at different stages of their cardio-metabolic disease is characterized by high throughput sequencing of the total fecal DNA. This will provide an enormous reservoir for the discovery of unsuspected metagenomic signatures that can represent predictive biomarkers and unsuspected new therapeutic targets for different disease phenotypes or stages. The accompanying challenge proposed here is a holistic metaproteomics approach to move beyond the genetic potential addressed by metagenomics and get closer to the real functionality of the gut microbiome by exploring the expression of metabolic and cellular pathways. Understanding how they are altered in disease can have a profound impact on patient treatment and disease prevention. This may open new avenues for reducing risk, including the discovery of new biomarkers, new targets and new therapeutic molecules. ProteoCardis will follow a step-by-step approach that we have already successfully experienced on a modest scale. First, in a well-controlled shotgun metaproteomics approach, and without any a priori assumption of the metabolic and/or cellular pathways that can accompany the disease, we will search for metaproteomic variables that are associated with CAD in 50 patients with acute event and 100 patients with chronic coronaropathy with or without congestive heart failure (CHF), compared to 50 healthy controls (CTRL) matched with BMI and sex. The change of protein signals before and one year after bariatric surgery (BS), an intervention known to reduce the cardiovascular risk, will also be examined (two observation times in the same 20 BS subjects). Finally, we will provide a set of relevant variables from those obtained in both contexts of aggravation (CAD vs. CTRL) and improvement (BS2 vs. BS1) of the clinical status. Then, to confirm the candidates and test their predictive value in high-risk patients, a multiplexed SRM (selected reaction monitoring) assay will be developed, targeting and precisely quantifying those proteins/peptides of interest. The validation process will be done by a recognized team in an independent manner, first on a subset of 20 CAD patients and 20 matched controls. ProteoCardis will subsequently examine the presence of these signals in 150 additional individuals at high risk of CAD (50 subjects with metabolic syndrome, 50 obese and 50 type 2 diabetic subjects) and association of these signals with any risk factors as well as complications or adverse cardiovascular outcomes of the patients over a four year period. The expected results have two levels of impact: immediate, since the whole experiment as a Standard Operational Procedure (SOP) will define new standards and share best practice among the scientific community for quantitative metaproteomics; longer-term, by pioneering a new map of knowledge linking together metaproteomic, metagenomic and clinical data related to CAD pathology, that will contribute to the opening of new avenues for maintaining or restoring health.