Learning Sequence Determinants of Protein: Protein Interaction Specificity with Sparse Graphical Models

  • Hetunandan Kamisetty
  • Bornika Ghosh
  • Christopher James Langmead
  • Chris Bailey-Kellogg
Conference paper

DOI: 10.1007/978-3-319-05269-4_10

Volume 8394 of the book series Lecture Notes in Computer Science (LNCS)
Cite this paper as:
Kamisetty H., Ghosh B., Langmead C.J., Bailey-Kellogg C. (2014) Learning Sequence Determinants of Protein: Protein Interaction Specificity with Sparse Graphical Models. In: Sharan R. (eds) Research in Computational Molecular Biology. RECOMB 2014. Lecture Notes in Computer Science, vol 8394. Springer, Cham

Abstract

In studying the strength and specificity of interaction between members of two protein families, key questions center on which pairs of possible partners actually interact, how well they interact, and why they interact while others do not. The advent of large-scale experimental studies of interactions between members of a target family and a diverse set of possible interaction partners offers the opportunity to address these questions. We develop here a method, DgSpi (Data-driven Graphical models of Specificity in Protein:protein Interactions), for learning and using graphical models that explicitly represent the amino acid basis for interaction specificity (why) and extend earlier classification-oriented approaches (which) to predict the \(\varDelta{G}\) of binding (how well). We demonstrate the effectiveness of our approach in analyzing and predicting interactions between a set of 82 PDZ recognition modules, against a panel of 217 possible peptide partners, based on data from MacBeath and colleagues. Our predicted \(\varDelta{G}\) values are highly predictive of the experimentally measured ones, reaching correlation coefficients of 0.69 in 10-fold cross-validation and 0.63 in leave-one-PDZ-out cross-validation. Furthermore, the model serves as a compact representation of amino acid constraints underlying the interactions, enabling protein-level \(\varDelta{G}\) predictions to be naturally understood in terms of residue-level constraints. Finally, the model, DgSpi readily enables the design of new interacting partners, and we demonstrate that designed ligands are novel and diverse.

Keywords

protein:protein interaction specificity \(\varDelta{G}\) prediction graphical model PDZ 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Hetunandan Kamisetty
    • 1
  • Bornika Ghosh
    • 2
  • Christopher James Langmead
    • 3
  • Chris Bailey-Kellogg
    • 2
  1. 1.Department of BiochemistryUniversity of WashingtonUSA
  2. 2.Department of Computer ScienceDartmouthUSA
  3. 3.School of Computer ScienceCarnegie Mellon UniversityUSA