Gabriele Martelloni ¹ Alessio Turchi ² Chiara Fallerini ¹ Andrea Degl'innocenti ¹ Margherita Baldassarri ¹ Gen-Covid Multicenter Study ³ Simona Olmi ^4* Simone Furini ⁵ Alessandra Renieri ¹

• ¹ Medical Genetics Unit, Azienda Ospedaliera Universitaria Senese, Siena, Tuscany, Italy
• ² Osservatorio Astrofisico di Arcetri (INAF), Florence, Tuscany, Italy
• ³ Med Biotech Hub and Competence Center, Department of Medical Biotechnology, University of Siena, Siena, Tuscany, Italy
• ⁴ National Research Council (CNR), Roma, Italy
• ⁵ Department of Electrical, Electronic and Information Engineering, School of Engineering, University of Bologna, Bologna, Emilia-Romagna, Italy

The impact of common and rare variants in COVID-19 host genetics has been widely studied. In particular, in Fallerini et al. (2022) common and rare variants were used to define an interpretable machine learning model for predicting COVID-19 severity. Firstly, variants were converted into sets of Boolean features, depending on the absence or the presence of variants in each gene. An ensemble of LASSO logistic regression models was used to identify the most informative Boolean features with respect to the genetic bases of severity. After that, the Boolean features, selected by these logistic models, were combined into an Integrated PolyGenic Score, the so called IPGS, which offers a very simple description of the contribution of host genetics in COVID-19 severity. IPGS leads to an accuracy of 55-60 % on different cohorts and, after a logistic regression with in input both IPGS and the age, it leads to an accuracy of 75%. The goal of this paper is to improve the previous results, using not only the most informative Boolean features with respect to the genetic bases of severity but also the information on the host organs involved in the disease. We generalize here the IPGS adding a statistical weight for each organ, through the transformation of Boolean features into "Boolean quantum features", inspired by the Quantum Mechanics. The organs coefficients were set via the application of the genetic algorithm PyGad and, after that, we defined two new Integrated PolyGenic Score (IPGS_1^ph and IPGS_2^ph ). By applying a logistic regression with both IPGS, IPGS_2^ph (or indifferently IPGS_1^ph ) and age as input, we reach an accuracy of 84-86%, thus improving the results previously shown in Fallerini et al. (2022) by a factor of 10%..