Combining features in a graphical model to predict protein binding sites

Abstract

Large efforts have been made in classifying residues as binding sites in proteins using machine learning methods. The prediction task can be translated into the computational challenge of assigning each residue the label binding site or non-binding site. Observational data comes from various possibly highly correlated sources. It includes the structure of the protein but not the structure of the complex. The model class of conditional random fields (CRFs) has previously successfully been used for protein binding site prediction. Here, a new CRF-approach is presented that models the dependencies of residues using a general graphical structure defined as a neighborhood graph and thus our model makes fewer independence assumptions on the labels than sequential labeling approaches. A novel node feature "change in free energy" is introduced into the model, which is then denoted by ΔF-CRF. Parameters are trained with an online large-margin algorithm. Using the standard feature class relative accessible surface area alone, the general graph-structure CRF already achieves higher prediction accuracy than the linear chain CRF of Li et al. ΔF-CRF performs significantly better on a large range of false positive rates than the support-vector-machine-based program PresCont of Zellner et al. on a homodimer set containing 128 chains. ΔF-CRF has a broader scope than PresCont since it is not constrained to protein subgroups and requires no multiple sequence alignment. The improvement is attributed to the advantageous combination of the novel node feature with the standard feature and to the adopted parameter training method.

Proteins and Their Interacting Partners: An Introduction to Protein–Ligand Binding Site Prediction Methods

“…Sequence-based methods that predict protein–ligand binding sites and their interacting ligand-binding site residues are those that use information from evolutionary conservation and/or sequence similarity of homologous proteins. These methods can be broadly categorised into methods that utilize machine learning (Multi-RELIEF [ 11 ], TargetS [ 12 ], LigandRF [ 13 ], and OMSL [ 14 ]), methods that utilize only position-specific scoring matrices or PSSMs (INTREPID [ 15 ], DISCERN [ 16 ], ConSurf [ 17 ], and ConFunc [ 18 ]) and graph-based methods such as Conditional Random Field (CRF) [ 19 ]. The advent of including machine learning-based strategies into sequence-based methods has resulted in improved method sensitivity.…”

Optical Versus Electronic Implementation of Probabilistic Graphical Inference and Experimental Device Demonstration Using Nonlinear Photonics

“…With such a large number of variables, the decomposition by the chain rule factorization is predominantly advantageous in terms of the number of operations saved. The popularity of PGI is largely due to the emergence of big data in diverse disciplines, from medicine to economics to social networks [22]- [24].…”

Utilizing knowledge base of amino acids structural neighborhoods to predict protein-protein interaction sites

“…A number of descriptors have been utilized for the purpose of PPI identification, such as hydrophobicity [ 5 ], energy of solvatation [ 6 ], propensity [ 5 ] or RASA (Relative Solvent Accessible Surface Area) [ 3 – 6 ], with RASA being especially popular [ 7 ]. As for machine learning approaches, the best performing methods utilize Support Vector Machines (SVM) [ 3 , 5 ], Neural networks [ 8 ], Decision trees [ 6 ] or Conditional Random Fields (CRF) [ 9 , 10 ].…”

“…The goal is to find the most probable labeling of hidden variables according to observations. Our approach was inspired by the CRF-based method presented by Dong et al [ 9 ] and Wierschin et al [ 10 ] where a protein is represented in a graph. In that representation, every amino acid corresponds to a node, and two nodes are connected by an edge if their corresponding amino acids are sufficiently close to each other.…”