A Probabilistic and Continuous Model of Protein Conformational Space for Template-Free Modeling

Abstract

One of the major challenges with protein template-free modeling is an efficient sampling algorithm that can explore a huge conformation space quickly. The popular fragment assembly method constructs a conformation by stringing together short fragments extracted from the Protein Data Base (PDB). The discrete nature of this method may limit generated conformations to a subspace in which the native fold does not belong. Another worry is that a protein with really new fold may contain some fragments not in the PDB. This article presents a probabilistic model of protein conformational space to overcome the above two limitations. This probabilistic model employs directional statistics to model the distribution of backbone angles and 2 nd -order Conditional Random Fields (CRFs) to describe sequenceangle relationship. Using this probabilistic model, we can sample protein conformations in a continuous space, as opposed to the widely used fragment assembly and lattice model methods that work in a discrete space. We show that when coupled with a simple energy function, this probabilistic method compares favorably with the fragment assembly method in the blind CASP8 evaluation, especially on alpha or small beta proteins. To our knowledge, this is the first probabilistic method that can search conformations in a continuous space and achieves favorable performance. Our method also generated three-dimensional (3D) models better than template-based methods for a couple of CASP8 hard targets. The method described in this article can also be applied to protein loop modeling, model refinement, and even RNA tertiary structure prediction.

Proposition of ontology for safety using cased based reasoning in industrial context

“…In case retrieval, similarity scoring (or distance-based) is commonly used, such as taxonomy tree approach and use of tables or rules [6]. Another possible approach is to adopt a mix of approaches for different types of attributes; [7] used five different similarity scoring functions. In this research, the similarity scoring approach makes use of a semantic network to calculate the similarity score between two precursors.…”

Collective Estimation of Multiple Bivariate Density Functions With Application to Angular-Sampling-Based Protein Loop Modeling

“…The estimated probability density function of the Ramachandran plot is referred to as the Ramachandran distribution, which has become a fundamental concept in various protein structure-related problems, such as structural model checking (Laskowski et al, 1993;Hooft et al, 1997;Davis et al, 2004), protein structure prediction (Rohl et al, 2004;Zhao et al, 2010), side chain rotamer library (Bhuyan and Gao, 2011;Shapovalov and Dunbrack, 2011), and empirical energy functions (Buck et al, 2006). Ramachandran distributions are known to be affected by the secondary structure (Hovmöller et al, 2002;Jha et al, 2005) and the amino acid type (Berkholz et al, 2009) of the residue from which φ and ψ angles are calculated, as well as the neighboring amino acids (Keskin et al, 2004;Lennox et al, 2009;Ting et al, 2010 neighbor-dependent Ramachandran distributions is difficult because when we focus on a specific amino acid while conditioning on the neighboring amino acids, the data are fractionated into groups each of which may contain only a small number of data points.…”

“…One successful approach for computational protein structure prediction is angular-samplingbased methods (Rohl et al, 2004;Bystroff et al, 2000;Tuffery and Derreumaux, 2005;Hamelryck et al, 2006;Sellers et al, 2008;Boomsma et al, 2008;Mandell et al, 2009;Ting et al, 2010;Lennox et al, 2010;Zhao et al, 2010;Stein and Kortemme, 2013;Maadooliat et al, 2013b;Källberg et al, 2014). Loop structures are irregular parts of proteins which play important roles in protein function, stability and folding (Fetrow, 1995).…”

“…Although most problems in the protein structure field depend on coordinate-based methods, backbone-angle-based methods have provided an attractive alternative approach in various protein structure-related problems, such as protein structure prediction (Simons et al, 1999;Hamelryck et al, 2006;Boomsma et al, 2008;Zhao et al, 2010), protein loop modeling (Ting et al, 2010), model quality assessment (Benkert et al, 2008;Gao et al, 2009;Archie and Karplus, 2009), prediction server ranking (Qiu et al, 2008;Maadooliat et al, 2013a), protein structure alignment (Miao et al, 2008;Challis and Schmidler, 2012), free energy function learning (Mu et al, 2005;Altis et al, 2008;Riccardi et al, 2009), and molecular dynamics simulation (Altis et al, 2007). In this paper, we focus on statistical modeling of the bivariate distribution of protein backbone angles.…”

A Science Driven Production Cyberinfrastructure—the Open Science Grid

“…Other smaller communities have specific publications that have benefited from usage of the OSG including bioinformatics [21,22], chemistry [23], mathematics [24], molecular dynamics [25], nanotechnology [26], neutrino physics [27,28] …”