Inferring interaction partners from protein sequences

Abstract

Specific protein-protein interactions are crucial in the cell, both to ensure the formation and stability of multi-protein complexes, and to enable signal transduction in various pathways. Functional interactions between proteins result in coevolution between the interaction partners. Hence, the sequences of interacting partners are correlated. Here we exploit these correlations to accurately identify which proteins are specific interaction partners from sequence data alone. Our general approach, which employs a pairwise maximum entropy model to infer direct couplings between residues, has been successfully used to predict the three-dimensional structures of proteins from sequences. Building on this approach, we introduce an iterative algorithm to predict specific interaction partners from among the members of two protein families. We assess the algorithm's performance on histidine kinases and response regulators from bacterial two-component signaling systems. The algorithm proves successful without any a priori knowledge of interaction partners, yielding a striking 0.93 true positive fraction on our complete dataset, and we uncover the origin of this surprising success. Finally, we discuss how our method could be used to predict novel protein-protein interactions. INTRODUCTIONMany key cellular processes are carried out by interacting proteins. For instance, transient protein-protein interactions determine signaling pathways, and their specificity ensures proper signal transduction. Hence, mapping specific protein-protein interactions is central to a systems-level understanding of cells, and has broad applications to areas such as drug targeting. High-throughput experimental methods have recently elucidated a substantial fraction of protein-protein interactions in a few model organisms [1], but experimental approaches remain challenging. Meanwhile, major progress in sequencing has led to an explosion of available sequence data. Can we exploit this abundant new sequence data to identify specific protein-protein interaction partners?Specific interactions between proteins imply evolutionary constraints on the interacting partners. For instance, mutation of a contact residue in one partner generically impairs binding, but may be compensated by a complementary mutation in the other partner. This co-evolution of interaction partners results in a correlation of their amino-acid sequences. Similar correlations exist within single proteins, between amino acids that are in contact in the folded protein. However, the simple fact of a correlation between residues in a multiple sequence alignment is only weakly predictive of a three-dimensional * anne-florence.bitbol@upmc.fr † ljc37@cam.ac.uk; L.J.C. and N.S.W. contributed equally to this work. ‡ wingreen@princeton.edu contact [2-4], as correlation can also stem from other effects such as phylogeny and indirect interactions. Fortunately, global statistical models provide a means to disentangle direct and indirect interactions [5][6][7]. In particular, the maximum entropy pri...

Applications of sequence coevolution in membrane protein biochemistry

“…To extend protein-protein interaction prediction to a proteomic level, new algorithms have been devised to systematically identify protein-protein interactions [95,96]. Both studies used TCS as a benchmark case because of their rich sequence information and known interaction pairing.…”

“…The best scoring test protein pairs are considered correct and then added to the training set and the process is repeated. This iterative approach is rather robust as it can be done with no training set [95] or without ever recalculating the interaction energy of the training set to throw out outlier matched pairs [96]. These algorithms were extended to subunits of ABC transporters [95] and to members of the tryptophan biosynthesis machinery [96] and they will likely be generally applicable to determine whether two proteins interact, as long as enough sequence information is available.…”

Coevolutive, Evolutive and Stochastic Information in Protein-Protein Interactions

“…That condition imposes distinct boundaries for the amount of PPIs amenable of resolution across definitions of the stochastic variables in terms of coupled and uncoupled amino acids. Indeed, the expectation value 15) for the fraction M −1 n of primary sequence matches among native-like MSA models mapped by S in eq. [14] decreases substantially with ω S meaning that ⟨ϵ⟩ S is systematically larger for physicallycoupled amino-acids at various resolutions δ I (Fig.…”

On generative models of T-cell receptor sequences

“…Those families are believed to have evolved over long time scales under a shared selective pressure that shapes their statistics. For such families, physics-inspired statistical inference methods have helped to predict contacts between amino-acids in the protein [11], define sectors of co-evolving residues [12], or find interaction partners [13]. Deep [14] and non-deep [15] machine learning approaches have also been successfully applied.…”