The pre-calculated interactions were obtained by performing Interolog-based computational approach on Triticum aestivum and Puccinia proteomes.

Step 1: The whole proteomes of T. aestivum and Puccinia species were BLAST searched against the known (standard) protein-protein interaction databases (DIP, BioGRID, HPIDB, IntAct, MINT, PHI-base, and STRING).

Step 2: 112 BLAST parameter combinations were randomly generated on the basis of sequence identity (30%, 40%, 50%, 60%), evalue (1e-10, 1e-50, 1e-05, 1e-04, 1e-20, 1e-30, 1e-25), and sequence coverage (40%, 50%, 60%, 80%).

Step 3: The combination with identity >= 30%, evalue <= 1e-04, and coverage >= 40% was considered as the best combination as the highest number of Puccinia effector proteins were identified in the interactions predicted using this combination.

We have also provided the respective interaction databases, in reference to which these interactions have been predicted.