BIOINFORMATICS ORIGINAL PAPER Structural bioinformatics PROFcon: novel prediction of long-range contacts

Citations in the corpus

• Deep Spatio-Temporal Architectures and Learning for Protein Structure Prediction (2012, methodology): “Rather, we focus on a minimal set of features commonly used in machine learning-based contact prediction [11, 2, 21, 7, 24, 5].”
• Knowledge-based machine learning methods for macromolecular 3D structure prediction (2016, background): “Many machine-learning methods have been developed for protein contact prediction in the past decades (Fariselli and Casadio, 1999; Gobel et al., 2004; Olmea and Valencia, 1997; Punta and Rost, 2005; Vendruscolo and Domany, 1998; Vullo et al., 2006).”
• Predicting protein contact map using evolutionary and physical constraints by integer programming (2013, methodology): “Many machine-learning methods have been developed for protein contact prediction in the past decades (Fariselli and Casadio, 1999; Göbel et al., 2004; Olmea and Valencia, 1997; Punta and Rost, 2005; Vendruscolo and Domany, 1998; Vullo et al., 2006).”
• HHsenser: exhaustive transitive profile search using HMM–HMM comparison (2006, background): “Examples are the prediction of secondary structure (1), solvent accessebility (2), disorder (3), transmembrane helices (4), intraprotein contacts (5), protein–protein interactions (6), subcellular localization (7), internal repeats (8), deleterious mutations (9) and conserved or subtype-specific fun”
• Deep architectures for protein contact map prediction (2012, methodology): “Machine learning approaches use methods such as neural networks (Fariselli et al., 2001; Punta and Rost, 2005; Shackelford and Karplus, 2007), recursive neural networks (Baldi and Pollastri, 2003; Vulloet al.”
• Natively unstructured regions in proteins identified from contact predictions (2007, methodology): “As negatives (wellstructured ), we chose PDB proteins that were used to test PROFcon, i.e. none of the proteins had been used for its development (Punta and Rost, 2005).”
• De Novo Structure Prediction of Globular Proteins Aided by Sequence Variation-Derived Contacts (2014, methodology): “Support Vector Machine and neural network based approaches) [10,11,12,13] to the ones based on.”
• Toward an accurate prediction of inter-residue distances in proteins using 2D recursive neural networks (2014, background): “Among protein structural features, contact maps have been proposed as an intermediate step in assembling the unknown protein 3D structure from its amino acid sequence [16-22].”
• Prediction of protein beta-residue contacts by Markov logic networks with grounding-specific weights (2009, background): “Sequence features consist of: (1) a 20-dimensional vector of amino acid composition; (2) a 3-dimensional vector of frequencies of secondary structure assignments; (3) the average solvent accessibility; (4) chain length, described by a unary-encoding 4-dimensional vector as in (Punta and Rost, 2005).”
• Optimal contact definition for reconstruction of Contact Maps (2010, methodology): “Furthermore they have been used as a proxy for 3-dimensional structure prediction by means of machine learning techniques in order to predict residue contacts from sequence information [14-18].”
• Modeling the language of life – Deep Learning Protein Sequences (2019, background): “Other objectives that used the same idea included the prediction of transmembrane regions [11-13], solvent accessibility [14], residue flexibility (B-values) [15, 16], inter-residue contacts [17] and protein disorder [15, 18-20].”
• SNAP: predict effect of non-synonymous polymorphisms on function (2007, background): “Similar findings have been reported by other studies focusing on the prediction of very different aspects of protein function (40,55,56).”