An investigation of phosphopeptide binding to SH2 domain

Semi-supervised prediction of SH2-peptide interactions from imbalanced high-throughput data

Src homology 2 (SH2) domains are the largest family of the peptide-recognition modules (PRMs) that bind to phosphotyrosine containing peptides. Knowledge about binding partners of SH2-domains is key for a deeper understanding of different cellular processes. Given the high binding specificity of SH2, in-silico ligand peptide prediction is of great interest. Currently however, only a few approaches have been published for the prediction of SH2-peptide interactions. Their main shortcomings range from limited coverage, to restrictive modeling assumptions (they are mainly based on position specific scoring matrices and do not take into consideration complex amino acids inter-dependencies) and high computational complexity. We propose a simple yet effective machine learning approach for a large set of known human SH2 domains. We used comprehensive data from micro-array and peptide-array experiments on 51 human SH2 domains. In order to deal with the high data imbalance problem and the high signal-to-noise ration, we casted the problem in a semi-supervised setting. We report competitive predictive performance w.r.t. state-of-the-art. Specifically we obtain 0.83 AUC ROC and 0.93 AUC PR in comparison to 0.71 AUC ROC and 0.87 AUC PR previously achieved by the position specific scoring matrices (PSSMs) based SMALI approach. Our work provides three main contributions. First, we showed that better models can be obtained when the information on the non-interacting peptides (negative examples) is also used. Second, we improve performance when considering high order correlations between the ligand positions employing regularization techniques to effectively avoid overfitting issues. Third, we developed an approach to tackle the data imbalance problem using a semi-supervised strategy. Finally, we performed a genome-wide prediction of human SH2-peptide binding, uncovering several findings of biological relevance. We make our models and genome-wide predictions, for all the 51 SH2-domains, freely available to the scientific community under the following URLs: http://www.bioinf.uni-freiburg.de/Software/SH2PepInt/SH2PepInt.tar.gz and http://www.bioinf.uni-freiburg.de/Software/SH2PepInt/Genome-wide-predictions.tar.gz, respectively.

Using genome-wide measurements for computational prediction of SH2-peptide interactions

Peptide-recognition modules (PRMs) are used throughout biology to mediate protein–protein interactions, and many PRMs are members of large protein domain families. Recent genome-wide measurements describe networks of peptide–PRM interactions. In these networks, very similar PRMs recognize distinct sets of peptides, raising the question of how peptide-recognition specificity is achieved using similar protein domains. The analysis of individual protein complex structures often gives answers that are not easily applicable to other members of the same PRM family. Bioinformatics-based approaches, one the other hand, may be difficult to interpret physically. Here we integrate structural information with a large, quantitative data set of SH2 domain–peptide interactions to study the physical origin of domain–peptide specificity. We develop an energy model, inspired by protein folding, based on interactions between the amino-acid positions in the domain and peptide. We use this model to successfully predict which SH2 domains and peptides interact and uncover the positions in each that are important for specificity. The energy model is general enough that it can be applied to other members of the SH2 family or to new peptides, and the cross-validation results suggest that these energy calculations will be useful for predicting binding interactions. It can also be adapted to study other PRM families, predict optimal peptides for a given SH2 domain, or study other biological interactions, e.g. protein–DNA interactions.

Genome-wide prediction of SH2 domain targets using structural information and the FoldX algorithm

Current experiments likely cover only a fraction of all protein-protein interactions. Here, we developed a method to predict SH2-mediated protein-protein interactions using the structure of SH2-phosphopeptide complexes and the FoldX algorithm. We show that our approach performs similarly to experimentally derived consensus sequences and substitution matrices at predicting known in vitro and in vivo targets of SH2 domains. We use our method to provide a set of high-confidence interactions for human SH2 domains with known structure filtered on secondary structure and phosphorylation state. We validated the predictions using literature-derived SH2 interactions and a probabilistic score obtained from a naive Bayes integration of information on coexpression, conservation of the interaction in other species, shared interaction partners, and functions. We show how our predictions lead to a new hypothesis for the role of SH2 domains in signaling.

Understanding the functional role of every protein in the cell is a long-standing goal of cellular biology. An important step in this direction is to discover how and when proteins interact inside the cell to accomplish their tasks. Many of the cellular functions depend on reversible protein modifications like phosphorylation. To sense these modifications, cells have protein domains capable of binding phosphorylated proteins such as the SH2 domain. In this work, we show that it is possible to use the three-dimensional structure of protein domains to predict its binding preferences. Using a computational tool called FoldX, we have predicted the binding specificity of several human SH2 domains. These predictions, based on the computational analysis of the 3-D structure, were shown to be of similar accuracy as those obtained from experimental binding assays. We show here that it is also possible to understand how a mutation changes the binding preference of protein binding domains, opening the way for better understanding of some disease causing mutations. The combination of this novel computational approach with other sources of information allowed us to provide a set of high-confidence novel interactions for the proteins here studied.

Calculation of absolute protein-ligand binding free energy from computer simulations

A general methodology for calculating the equilibrium binding constant of a flexible ligand to a protein receptor is formulated on the basis of potentials of mean force. The overall process is decomposed into several stages that can be computed separately: the free ligand in the bulk is first restrained into the conformation it adopts in the bound state, position, and orientation by applying biasing potentials, then it is translated into the binding site, where it is released completely. The conformational restraining potential is based on the root-mean-square deviation of the peptide coordinates relative to its average conformation in the bound complex. Free energy contributions from each stage are calculated by means of free energy perturbation potential of mean force techniques by using appropriate order parameters. The present approach avoids the need to decouple the ligand from its surrounding (bulk solvent and receptor protein) as is traditionally performed in the double-decoupling scheme. It is believed that the present formulation will be particularly useful when the solvation free energy of the ligand is very large. As an application, the equilibrium binding constant of the phosphotyrosine peptide pYEEI to the Src homology 2 domain of human Lck has been calculated. The results are in good agreement with experimental values.