The annotation of full zinc proteomes

Abstract

We obtained an extended functional annotation of zinc proteins using a combination of bioinformatic methods. This work was performed using a number of available predicted zinc proteomes of various representative organisms, leading to the almost complete annotation of, among others, the predicted human zinc proteome. The computational tools exploited included sequence-based and, when possible, structure-based functional predictions. We assigned a hypothetical function to 74% of the 1,472 sequences analyzed that lacked annotation in the starting dataset. We also added new functional categories, not described in the reference dataset, such as ubiquitin binding and DNA replication. As a general conclusion, we can state that the quality of each functional prediction parallels the amount of information for the sequence analyzed: the larger the amount of information, the more detailed and reliable is the proposed functional prediction. Among the findings, we have propose a zinc binding site for archaeal zinc-importing proteins. Furthermore, we propose two groups of transcriptional regulators that are involved in fatty acid metabolism.