A genetic features and gene interaction study for identifying the genes that cause hereditary spherocytosis

<b>Objective:</b> Hereditary spherocytosis (HS) is a hemolytic disorder characterized by the presence of spherical-shaped red blood cells on the peripheral blood smear. Non-dominant HS cases are due to de novo mutations of the type associated with dominant inheritance or recessive genes. This study is aimed to identify HS-related biological mechanisms and predicting HS candidate genes. <b>Methods:</b> We searched the known HS-related genes from the public databases. By analyzing the gene ontology (GO) and biological pathway of these genes, we extracted the optimal features to encode HS genes. Based on them, we predicted the HS-related genes from genes of whole genomes using the Random Forest classification. We used the gene interaction networks analysis to further identify the core regulatory genes that were related to HS. <b>Results:</b> Forty-one known HS-related genes were found out and encoded. Three hundred and sixty-seven GO terms and ten biological pathway terms were identified as the optimal features for prediction. We subsequently predicted 150 novel HS-related genes and identified the core regulatory genes in the interaction network of predicted and known genes. These features and genes that we identified could complement the genetic features of HS.