Construction and analysis of an interologous protein-protein interaction network of Camellia sinensisleaf (TeaLIPIN) from RNA-Seq datasets

Worldwide, tea (Camellia sinensis) is the most consumed beverage primarily due to the taste, flavour and aroma of its newly formed leaves; and has been used as an important ingredient in several traditional medicinal systems because of its antioxidant properties. For this medicinally and commercially important plant, design principles of gene-regulatory and protein-protein interaction networks at sub-cellular level are largely un-characterized mainly because neither its reference assembly nor annotated genome is available though plenty of its transcriptomes have been studied. In this work, we report a tea leaf interologous PPI network (TeaLIPIN) consisting of 11,208 nodes and 1,97,820 interactions. A reference transcriptome assembly was first developed from all the 44 samples of 6 publicly available leaf transcriptomes (1,567,288,290 raw reads). By inferring the high-confidence interactions among potential proteins coded by these transcripts using known experimental information about PPIs in 14 plants, an interologous PPI network was constructed and its modular architecture was explored. Comparing this PPI network with 10,000 realisations of corresponding random networks (Erd[o]s-Renyi models) and examining over three network centrality metrics, we predict 2,931 bottleneck proteins (having p-values <0.01). 270 of these are deduced to have transcription factor domains by developing the HMM models of known plant TFs. Final transcripts were also mapped to the draft tea genome in order to search the probable loci of their origin. We believe that the proposed novel methodology can easily be adopted to develop and explore the PPI interactomes in other plant species by making use of the available transcriptomic data.