Analysis and prediction of albinism co-phenotypes based on co-occurrence network

Objective To analyze the phenotypic characteristics of albinism and their correlations.

Methods Literatures related to albinism in PubMed database before June 10, 2023 were obtained, and phenotypic entities were extracted from their abstract texts to construct a phenotypic co-occurrence network. GePhi and VOSviewer software were used to analyze the overall characteristics and phenotypic clustering of the network. The Apriori algorithm was used to mine the association rules between phenotypes. The AA index was used for link prediction, predicting possible combinations of albino phenotypes.

Results The phenotypic co-occurrence network of albinism had a small-world characteristic, and nystagmus, hypopigmentation of skin and foveal dysplasia of retina were the main phenotypes. Phenotypic abnormalities of albinism were mainly divided into five categories, including visual system abnormalities, immune system abnormalities, skin and hair system abnormalities, nervous system abnormalities, and respiratory as well as digestive system abnormalities. A variety of ocular abnormalities were associated with nystagmus. The combination of abnormal phenotypes that may occur in albinism included pneumonia with hypopigmentation of the skin, foveal dysplasia of the retina and atrophy of the optic nerve.

Conclusion Predicting and analyzing the phenotypic characteristics of albinism and the associative patterns among phenotypes, as well as the potential co-occurrence of phenotypes in albinism patients can provide valuable references for identifying research directions for albinism and for the recognition, diagnosis, and prediction of albinism development.