RESUMO
With the development of deep sequencing, a significant proportion of mutations already listed in studies have inconclusive pathogenicity. We aim to establish the proportion of cases in which familial studies are possible and cosegregation analysis is informative. We also compare cosegregation analysis with in silico software and a proposed pathogenicity score. 204 consecutive positive tests were reviewed. 4 different in silico software programs were used. Spaendonck-Zwarts' pathogenicity score was also calculated. A total of 73 of the missense variants could be classified by the score as being likely or definitively pathogenic. A high percentage of nonsense variants were found in desmosomal genes and missense variants in sarcomeric genes. 36.3% of the missense variants in our cohort classified as very likely or definitively pathogenic were novel. Cosegregation analysis was positive in 19.5% and could be discarded in 15.6%. There was a significant discrepancy between the in silico tools used in the setting of inherited heart disease. Multiparametric scoring systems which include cosegregation and functional studies seem to perform better than individual prediction software.
