Use of a structural alphabet for analysis of short loops connecting repetitive structures

Fourrier, Laurent; Benros, Cristina; Brevern, Alexandre G. de

doi:10.1186/1471-2105-5-58

Cited by 55 publications

(17 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It corroborates previous observations that have already noted a high divergence of this method with every other SSAMS. [14] SSAMS that assigns less α-helix (mean value = 36.74%) assigns more β-strand (mean value = 22.25%). Analysis of coil frequency gives two major clusters above and below 17%.…”

Section: Descriptionmentioning

confidence: 99%

Protein beta-turn assignments

Bornot¹,

Brevern²

2006

Bioinformation

View full text Add to dashboard Cite

Abstract:A classical way to analyze protein 3D structures or models is to investigate their secondary structures. Their predictions are also widely used as a help to build new 3D models. Thus, hundreds of prediction methods have been proposed. Nonetheless before predicting, secondary structure assignment is required even if not trivial. Therefore numerous but diverging assignment methods have been developed. β-turns constitute the third most important secondary structures. However, no analysis to compare the β-turn distributions according to different secondary structure assignment methods has ever been done. We propose in this paper to analyze and evaluate the results of such a comparison. We highlight some important divergence that could have important consequence for the analysis and prediction of β-turns.

show abstract

Section: Descriptionmentioning

confidence: 99%

Protein beta-turn assignments

Bornot¹,

Brevern²

2006

Bioinformation

View full text Add to dashboard Cite

show abstract

“…Our earlier work on PBs have shown that PBs are effective in describing and predicting conformations of long fragments (Benros, et al, 2006; Benros, et al, 2009; Bornot, et al, 2009; de Brevern, et al, 2007; de Brevern and Hazout, 2001; de Brevern and Hazout, 2003; de Brevern, et al, 2002) and short loops (Fourrier, et al, 2004; Tyagi, et al, 2009; Tyagi, et al, 2009), analyzing protein contacts (Faure, et al, 2008), in building a transmembrane protein (de Brevern, 2005; de Brevern, et al, 2009), and in defining a reduced amino acid alphabet to aid design of mutations (Etchebest, et al, 2007). This reduced amino acid alphabet was recently proved suitable for predicting protein families or sub-families and secretory proteins of P. falciparum (Zuo and Li, 2009; Zuo and Li, 2009).…”

Section: Protein Blocksmentioning

confidence: 99%

“…A website, LocPred (http://www.dsimb.inserm.fr/~debrevern/LOCPRED/), which includes most of the tools developed so far, is available to perform these predictions (de Brevern, et al, 2004). Predictions were also performed with the SWs (de Brevern, et al, 2007; de Brevern, et al, 2002) and specifically for short loops (Fourrier, et al, 2004; Tyagi, et al, 2009). …”

Section: Applicationsmentioning

confidence: 99%

A short survey on protein blocks

et al. 2010

View full text Add to dashboard Cite

Protein structures are classically described in terms of secondary structures. Even if the regular secondary structures have relevant physical meaning, their recognition from atomic coordinates has some important limitations such as uncertainties in the assignment of boundaries of helical and β-strand regions. Further, on an average about 50% of all residues are assigned to an irregular state, i.e., the coil. Thus different research teams have focused on abstracting conformation of protein backbone in the localized short stretches. Using different geometric measures, local stretches in protein structures are clustered in a chosen number of states. A prototype representative of the local structures in each cluster is generally defined. These libraries of local structures prototypes are named as “structural alphabets”. We have developed a structural alphabet, named Protein Blocks, not only to approximate the protein structure, but also to predict them from sequence. Since its development, we and other teams have explored numerous new research fields using this structural alphabet. We review here some of the most interesting applications.

show abstract

“…The template-free [74-81, 83, 84] methods reported accuracies from 82 to 87%, a substantial improvement over the fourth-generation accuracy of 80%. The theoretical 3-label accuracy limit is 90-95% [1,30,72,82,85,86], based on the rate at which different programs such as DSSP and Stride [87] assign the same secondary structure to experimental coordinates.…”

Section: Introductionmentioning

confidence: 99%

Multifaceted analysis of training and testing convolutional neural networks for protein secondary structure prediction

2020

View full text Add to dashboard Cite

Protein secondary structure prediction remains a vital topic with broad applications. Due to lack of a widely accepted standard in secondary structure predictor evaluation, a fair comparison of predictors is challenging. A detailed examination of factors that contribute to higher accuracy is also lacking. In this paper, we present: (1) new test sets, Test2018, Test2019, and Test2018-2019, consisting of proteins from structures released in 2018 and 2019 with less than 25% identity to any protein published before 2018; (2) a 4-layer convolutional neural network, SecNet, with an input window of ±14 amino acids which was trained on proteins �25% identical to proteins in Test2018 and the commonly used CB513 test set; (3) an additional test set that shares no homologous domains with the training set proteins, according to the Evolutionary Classification of Proteins (ECOD) database; (4) a detailed ablation study where we reverse one algorithmic choice at a time in SecNet and evaluate the effect on the prediction accuracy; (5) new 4-and 5-label prediction alphabets that may be more practical for tertiary structure prediction methods. The 3-label accuracy (helix, sheet, coil) of the leading predictors on both Test2018 and CB513 is 81-82%, while Sec-Net's accuracy is 84% for both sets. Accuracy on the non-homologous ECOD set is only 0.6 points (83.9%) lower than the results on the Test2018-2019 set (84.5%). The ablation study of features, neural network architecture, and training hyper-parameters suggests the best accuracy results are achieved with good choices for each of them while the neural network architecture is not as critical as long as it is not too simple. Protocols for generating and using unbiased test, validation, and training sets are provided. Our data sets, including input features and assigned labels, and SecNet software including third-party dependencies and databases, are downloadable from dunbrack.fccc.edu/ss and github.com/sh-maxim/ss.

show abstract

Use of a structural alphabet for analysis of short loops connecting repetitive structures

Cited by 55 publications

References 45 publications

Protein beta-turn assignments

Protein beta-turn assignments

A short survey on protein blocks

Multifaceted analysis of training and testing convolutional neural networks for protein secondary structure prediction

Contact Info

Product

Resources

About