Computational prediction of protein–protein interactions’ network in Arabidopsis thaliana

Abstract

The study of protein–protein interactions (PPIs) has been a major factor in understanding the function of proteins. The development of diverse methodologies currently facilitates the identification of novel or uncharacterized PPIs. Despite advancements in other species, there is no complete interactome map for plants. Therefore, sketching an interactome map based on the interolog method using a large number of species will further our understanding in this field. We employed interolog to develop the Arabidopsis PPI network (PPIN) in the current investigation. We used data from 273 species to construct our PPIN by collecting experimentally reported PPIs from databases (IntAct, BioGrid, Mint, and DIP) and then using InParanoid to identify the orthologous proteins. The final Arabidopsis predicted PPIN consisted of 526,367 interactions between 10 and 105 proteins. The final PPIN was constructed based on the selection of reliable data sources, and a threshold was applied to filter predicted interactions to reach 289,909 interactions. We believe this predicted PPIN can contribute to ongoing research and provide an excellent opportunity for plant interactomic studies. In addition, the availability of Arabidopsis PPIN increases knowledge at the protein level, facilitates a better understanding of signal transduction pathways, and enables the identification of new proteins involved in unique processes.