Bayesian Machine Learning Techniques for revealing complex interactions among genetic and clinical factors in association with extra-intestinal Manifestations in IBD patients.

The objective of the study is to assess the predictive performance of three different techniques as classifiers for extra-intestinal manifestations in 152 patients with Crohn's disease. Naïve Bayes, Bayesian Additive Regression Trees and Bayesian Networks implemented using a Greedy Thick Thinning algorithm for learning dependencies among variables and EM algorithm for learning conditional probabilities associated to each variable are taken into account. Three sets of variables were considered: (i) disease characteristics: presentation, behavior and location (ii) risk factors: age, gender, smoke and familiarity and (iii) genetic polymorphisms of the NOD2, CD14, TNFA, IL12B, and IL1RN genes, whose involvement in Crohn's disease is known or suspected. Extra-intestinal manifestations occurred in 75 patients. Bayesian Networks achieved accuracy of 82% when considering only clinical factors and 89% when considering also genetic information, outperforming the other techniques. CD14 has a small predicting capability. Adding TNFA, IL12B to the 3020insC NOD2 variant improved the accuracy.

X-linked Alport syndrome: natural history in 195 families and genotype- phenotype correlations in males.

Alport syndrome (AS) is a type IV collagen hereditary disease characterized by the association of progressive hematuric nephritis, hearing loss, and, frequently, ocular changes. Mutations in the COL4A5 collagen gene are responsible for the more common X-linked dominant form of the disease. Considerable allelic heterogeneity has been observed. A "European Community Alport Syndrome Concerted Action" has been established to delineate accurately the AS phenotype and to determine genotype-phenotype correlations in a large number of families. Data concerning 329 families, 250 of them with an X-linked transmission, were collected. Characteristics of the 401 male patients belonging to the 195 families with COL4A5 mutation are presented. All male patients were hematuric, and the rate of progression to end-stage renal failure and deafness was mutation-dependent. Large deletions, non-sense mutations, or small mutations changing the reading frame conferred to affected male patients a 90% probability of developing end-stage renal failure before 30 yr of age, whereas the same risk was of 50 and 70%, respectively, in patients with missense or splice site mutation. The risk of developing hearing loss before 30 yr of age was approximately 60% in patients with missense mutations, contrary to 90% for the other types of mutations. The natural history of X-linked AS and correlations with COL4A5 mutations have been established in a large cohort of male patients. These data could be used for further evaluation of therapeutic approaches.