Omics approaches has improved our knowledge in the field of many genetic disorders. Despite this technological advancement, there are pathologies in which genotype–phenotype correlation and the ability to predict clinical outcomes still remain unsatisfactory. In people with a Marfan-like habitus there is a defect at connective tissue level and a great number of cardiovascular, skeletal, ophthalmological and dermatological abnormalities can be observed. Although the identification of pathogenic genetic variants contributes to the diagnostic process, its value to the prediction of clinical outcomes is still limited. With this project we aim to improve the genotypephenotype correlation in the field of Marfan-like spectrum by widening the amount of genetic information used and adopting an innovative statistical approach. We aim to use clinical exome sequencing data obtained from a cohort of patients with Marfan-like habitus and estimate the effects that both rare and common variants exert on phenotype. Finally, we aim to combine all the statistically significant effects in a unique inclusive statistical model using structural equation models and machine learning approaches