👤 Alessio Gerussi

Cholestatic liver diseases are a heterogeneous group of conditions that can remain unexplained despite a comprehensive diagnostic assessment. Genetic disorders may underlie many of these unexplained adult-onset cholestasis cases. However, genetic testing in adults has been focused on genes linked to progressive familial intrahepatic cholestasis (PFIC). This study evaluated the diagnostic utility of whole exome sequencing (WES) by targeting a broader set of genes beyond PFIC genes. Adults with unexplained cholestatic liver disease from one tertiary center underwent WES. Pathogenic and rare damaging variants in candidate cholestatic and liver disease genes were prioritized, and genotype-phenotype correlations were conducted. Twenty-one patients with three distinct cholestatic phenotypes (recurrent lithiasis, intrahepatic cholestasis, and primary sclerosing cholangitis with unusual features) were included. WES yielded a genetic diagnosis of inherited cholestatic or liver disorder mimicking the cholestatic phenotype in 5 cases (23.8%). ABCB4 was the causative gene in 2 cases (40.0%), whereas genes outside the PFIC spectrum (ABCC2, PPOX, APOB) accounted for the other 3 (60.0%). This study highlights the value of WES in the diagnostic workup of adult-onset cholestatic liver disease and expands our understanding of its genetic landscape, paving the way for larger-scale studies. Show less

The application of Machine Learning (ML) to genetic individual-level data represents a foreseeable advancement for the field, which is still in its infancy. Here, we aimed to evaluate the feasibility and accuracy of an ML-based model for disease risk prediction applied to Primary Biliary Cholangitis (PBC). Genome-wide significant variants identified in subjects of European ancestry in the recently released second international meta-analysis of GWAS in PBC were used as input data. Quality-checked, individual genomic data from two Italian cohorts were used. The ML included the following steps: import of genotype and phenotype data, genetic variant selection, supervised classification of PBC by genotype, generation of "if-then" rules for disease prediction by logic learning machine (LLM), and model validation in a different cohort. The training cohort included 1345 individuals: 444 were PBC cases and 901 were healthy controls. After pre-processing, 41,899 variants entered the analysis. Several configurations of parameters related to feature selection were simulated. The best LLM model reached an Accuracy of 71.7%, a Matthews correlation coefficient of 0.29, a Youden's value of 0.21, a Sensitivity of 0.28, a Specificity of 0.93, a Positive Predictive Value of 0.66, and a Negative Predictive Value of 0.72. Thirty-eight rules were generated. The rule with the highest covering (19.14) included the following genes: RIN3, KANSL1, TIMMDC1, TNPO3. The validation cohort included 834 individuals: 255 cases and 579 controls. By applying the ruleset derived in the training cohort, the Area under the Curve of the model was 0.73. This study represents the first illustration of an ML model applied to common variants associated with PBC. Our approach is computationally feasible, leverages individual-level data to generate intelligible rules, and can be used for disease prediction in at-risk individuals. Show less