Transcriptome-driven Health-status Transversal-predictor Analysis (THTA) using the PBMC transcriptome for health, food, microbiome and disease markers for understanding the background and development of lifestyle diseases

We developed a novel machine-learning artificial intelligence (AI) approach to predict general health and food-intake parameters named Transcriptome-driven Health-status Transversal-predictor Analysis (THTA) with relevance for diabesity markers based on a mathematics-driven and non-transcriptomic biomarker driven approach. The prediction was based on values from food consumption, dietary lipids and their bioactive metabolites, peripheral blood mononuclear cells (PBMC) mRNA-based transcriptome signatures, magnetic resonance imaging (MRI), energy metabolism measurements, microbiome analyses, and baseline clinical parameters in a cohort of 72 subjects. Our novel machine learning approach included transcriptome data from PBMCs as a “one-method” approach to predict 77 general health-status markers for broad stratification of the diabesity phenotype, which are usually necessitating measurements using 16 different methods. The PBMC transcriptome was used to determine these selected 77 basic and background health-status markers in a transversal-predictor establishment group with very high accuracy (Pearson correlations are r = 0,94 ranging from 0,88 to 0,98). These collected variables offer a valuable indication to identify which individual factor(s) are mainly targeting diabesity. Based on the “establishment group“ prediction approach a further “confirmation group” prediction approach was performed with a predictive potential for these 77 variables of r = 0,62 (ranging from 0,30 to 0,99). This “one-method” approach allows monitoring of a large number of health-status variables with relevance for diabesity simultaneously and may enable monitoring of therapeutic and preventive strategies. In summary, this novel technique based on PBMC transcriptomics from human blood offers prediction of a large range of health-related markers, which independently would be obtained in different clinical / research centres at a much higher price.