Reversing the pipeline: a ‘human-first’ multi-omics approach to cardiovascular discovery

Figures

• 

• 

  Fig. 1. Human cardiovascular tissue multi-omics pipeline for mechanistic and translational discovery. The figure illustrates a structured human-first workflow for cardiovascular multi-omics studies. The pipeline begins with clinical cohort design, including disease phenotyping, matched comparison groups, imaging, medications, comorbidities, longitudinal outcomes and standardized metadata collection. Human cardiovascular tissues are then obtained from endomyocardial biopsies, surgical specimens, explanted hearts, donor hearts, autopsy material, vascular tissue or valvular tissue. Each sample undergoes pathology review and quality control, including anatomical sampling-site documentation, histology, ischemic time, fixation or freezing method, RNA integrity, proteomic quality metrics and assessment of tissue heterogeneity. Molecular profiling may include bulk transcriptomics, single-cell or single-nucleus RNA sequencing, spatial transcriptomics, spatial proteomics, proteomics, metabolomics, epigenomics and targetsed validation assays. Computational integration then harmonizes multilayered datasets using batch correction, cell-state annotation, spatial mapping, network inference, machine…

• 

  Fig. 2. Closing the translational failure gap in cardiovascular research. The figure contrasts the conventional animal-first translational pipeline with the proposed human-first iterative framework. In the traditional model, hypotheses are generated and tested primarily in animal models, leading to preclinical target nomination before clinical trials. Although this approach remains essential for mechanistic testing, it is limited by species-specific physiology, simplified disease modeling and an incomplete representation of human cardiovascular heterogeneity, which contribute to high clinical attrition. The human-first framework begins with direct molecular interrogation of diseased human cardiovascular tissues, followed by multi-omics profiling to identify candidate mechanisms present in the clinical disease state. These candidates are then tested in perturbable validation systems, including organoids, engineered tissues and animal models, to establish causality and therapeutic tractability. Validated mechanisms can subsequently be translated into circulating biomarkers, therapeutic targets, patient stratification tools and clinical studies. The loop from clinical studies back to human…