A preliminary analysis of gene pleiotropy estimated from protein sequences.

Biologists have long recognized the importance of gene pleiotropy, that is, single genes affect multiple traits, which is one of the most commonly observed attributes of genes. Yet the extent of gene pleiotropy has been seriously under-explored. Theoretically, Fisher's model assumed a universal pleiotropy, that is, a mutation can potentially affect all phenotypic traits. On the other hand, experimental assays of a gene usually showed a few distinct phenotypes. Our recent work provides a new approach by estimating the degree of pleiotropy effectively from the phylogenetic sequence analysis. In this article, we estimated the effective gene pleiotropy for 321 vertebrate genes, and found that a gene typically affects 6-7 molecular phenotypes that correspond to the components of organismal fitness, respectively. The positive correlation of gene pleiotropy with the number of Gene Ontology biological processes, as well as the expression broadness provides a biological basis for the sequence-based estimation of gene pleiotropy. On the other hand, the degree of gene pleiotropy has been restricted to a digital number of molecular phenotypes, indicating that some cautions are needed for theoretical analysis of gene pleiotropy based on the assumption of universal pleiotropy.