Decision trees for early prediction of inadequate immune response to coronavirus infections: a pilot study on COVID-19

Fabio Pisano; Barbara Cannas; Alessandra Fanni; Manuela Pasella; Beatrice Canetto; Sabrina Rita Giglio; Stefano Mocci; Luchino Chessa; Andrea Perra; Roberto Littera

doi:10.3389/fmed.2023.1230733

Decision trees for early prediction of inadequate immune response to coronavirus infections: a pilot study on COVID-19

Front Med (Lausanne). 2023 Aug 2:10:1230733. doi: 10.3389/fmed.2023.1230733. eCollection 2023.

Authors

Fabio Pisano¹, Barbara Cannas¹, Alessandra Fanni¹, Manuela Pasella¹, Beatrice Canetto², Sabrina Rita Giglio^{3

4

5

6}, Stefano Mocci^{3

6}, Luchino Chessa^{4

7

8}, Andrea Perra^{4

9}, Roberto Littera^{4

5}

Affiliations

¹ Department of Electrical and Electronic Engineering, University of Cagliari, Cagliari, Italy.
² BithiaTec Technologies, Elmas, Italy.
³ Medical Genetics, Department of Medical Sciences and Public Health, University of Cagliari, Cagliari, Italy.
⁴ AART-ODV (Association for the Advancement of Research on Transplantation), Cagliari, Italy.
⁵ Medical Genetics, R. Binaghi Hospital, Local Public Health and Social Care Unit (ASSL) of Cagliari, Cagliari, Italy.
⁶ Centre for Research University Services (CeSAR, Centro Servizi di Ateneo per la Ricerca), University of Cagliari, Cagliari, Monserrato, Italy.
⁷ Department of Medical Sciences and Public Health, University of Cagliari, Cagliari, Italy.
⁸ Liver Unit, Department of Internal Medicine, University Hospital of Cagliari, Cagliari, Italy.
⁹ Unit of Oncology and Molecular Pathology, Department of Biomedical Sciences, University of Cagliari, Cagliari, Italy.

Abstract

Introduction: Few artificial intelligence models exist to predict severe forms of COVID-19. Most rely on post-infection laboratory data, hindering early treatment for high-risk individuals.

Methods: This study developed a machine learning model to predict inherent risk of severe symptoms after contracting SARS-CoV-2. Using a Decision Tree trained on 153 Alpha variant patients, demographic, clinical and immunogenetic markers were considered. Model performance was assessed on Alpha and Delta variant datasets. Key risk factors included age, gender, absence of KIR2DS2 gene (alone or with HLA-C C1 group alleles), presence of 14-bp polymorphism in HLA-G gene, presence of KIR2DS5 gene, and presence of KIR telomeric region A/A.

Results: The model achieved 83.01% accuracy for Alpha variant and 78.57% for Delta variant, with True Positive Rates of 80.82 and 77.78%, and True Negative Rates of 85.00% and 79.17%, respectively. The model showed high sensitivity in identifying individuals at risk.

Discussion: The present study demonstrates the potential of AI algorithms, combined with demographic, epidemiologic, and immunogenetic data, in identifying individuals at high risk of severe COVID-19 and facilitating early treatment. Further studies are required for routine clinical integration.

Keywords: COVID-19; SARS-CoV-2; artificial intelligence; decision trees; disease severity; immunogenetic background.