Reducing dimensionality in remote homology detection using predicted contact maps

Bedoya, Oscar; Tischer, Irene

doi:10.1016/j.compbiomed.2015.01.020

Cited by 6 publications

(9 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is also expected that each superfamily has some submatrices of the contact maps and some distributions that are specific and different from the rest of the superfamilies. 6 can be observed in a beta-sheet. In addition, model m 9 is a submatrix that is usually observed in an anti-parallel betasheet.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Remote-3DD: un nuevo método para la detección de homólogos remotos que usa propiedades fisicoquímicas

Bedoya

2017

ingycompe

View full text Add to dashboard Cite

ResumenEn este artículo se presenta un nuevo método para la detección de homólogos remotos, llamado remote-3DD, que combina mapas de contacto predichos y una distribución de los valores en las matrices de interacción. Los mapas de contacto predichos son una aproximación de la forma 3D de proteína que se puede obtener a partir de su estructura primaria. Por su parte, una matriz de interacción permite representar una proteína a partir de las propiedades fisicoquímicas de los aminoácidos que la conforman. Remote-3DD se propone como una estrategia para mejorar la exactitud del método remote-C3D en el cual se utilizan solamente mapas de contacto. La hipótesis que se plantea en este artículo es que se puede mejorar la exactitud del método remote-C3D al incorporar las distribuciones de la matriz de interacción. Los resultados de las pruebas muestran que el método remote-3DD alcanza una exactitud mayor que los métodos basados en composición y en algunos casos una exactitud comparable con los métodos basados en perfiles. Además, las pruebas permiten demostrar que el método remote-3DD, en general, presenta exactitudes mayores que el método remote-C3D cuando se utiliza la misma cantidad de modelos y tamaños de submatrices. Palabras clave: Bioinformática, clasificadores, conjunto de datos SCOP, homólogos remotos, propiedades fisicoquímicas. AbstractIn this paper, we present a new method for remote homology detection called remote-3DD. The proposed method combines predicted contact maps and distributions of the interaction matrices. Predicted contact maps approximate the 3D shape of a protein based on its primary structure. On the other hand, an interaction matrix allows representing a protein by using the physicochemical properties of its amino acids. The remote-3DD method is proposed as a strategy to improve the accuracy of the remote-C3D method, which uses contact maps alone. In this paper, we hypothesize that we can improve the accuracy of the remote-C3D method by including physicochemical properties. The results show that the accuracy of the remote-3DD method is higher than the accuracy of the composition-based methods and in some cases comparable with the accuracy of the profilebased methods. In addition, the results also show that the remote-3DD method achieves higher accuracy values than the remote-C3D method when considering the same number of models and size of submatrices.

show abstract

Section: Resultsmentioning

confidence: 99%

“…This method is proposed as a strategy for improving the accuracy (i.e., ROC score) of the remote-C3D method (6), which uses contact maps alone. We hypothesize that the accuracy of the remote-C3D method can be improved by incorporating the distributions of the interaction matrices.…”

Section: Methodsmentioning

confidence: 99%

Remote-3DD: un nuevo método para la detección de homólogos remotos que usa propiedades fisicoquímicas

Bedoya

2017

ingycompe

View full text Add to dashboard Cite

show abstract

“…According to Russell et al (1997) [17] proteins that are able to perform the same function, either a remote homologue or analogue, can show less than 50 % of sequence similarity due to shared active sites or conserved domains; the formal definition of remote homology is protein sequences that share an identity percentage of less than 25 % [18]. mhpC gene is already annotated in R. eutropha H16 genome as aminoacrylate hydrolase and located in chromosome 2 while all PHA-related genes are located in chromosome 1 in R. eutropha H16 [11]; for those reasons, it was not included in further analysis as a potential PhaG analogue or remote homologue.…”

Section: In Silico Analysis and Phag Homologues Characterisationmentioning

confidence: 99%

In silico analysis of phag-like protein in Ralstonia Euthropa H16, potentially involved in polyhydroxyalkanoates synthesis

Acosta

Restrepo

2019

Rev. politec.

View full text Add to dashboard Cite

Polyhydroxyalkanoates (PHA) are synthesised by bacteria as carbon storage material. The protein PhaG directs carbon from non-related carbon sources such as glycerol, metabolised through fatty acid de novo synthesis (FAS) pathway, with PHA synthesis. The gene that codifies for this protein has not yet been found in the genome of Ralstonia eutropha H16, a model organism. By bioinformatic comparison to already known PhaG proteins, a PhaG-like protein was found codified by gene H16_A0147 and presence of the gene was preliminary confirmed by PCR. This is the first study that shows the presence and characteristics of a PhaG-like protein in R. eutropha H16 and represents the first step for the identification of a connection between FAS and PHA pathways in this model bacterium. Further gene deletion and enzymatic activity studies are necessary to confirm this potential relationship, which could improve industrial PHA production and utilisation of agro-industrial residues such as glycerol.

show abstract

“…Its binary form is known as a contact map. Lots of information can be directly obtained from them, such as secondary structure, 1 motif, 2 and interaction types 3 concerning the kinds of residues. Besides, conserved patterns and structure motifs can be found [1] by analyzing the distance or contact map.…”

Section: Introductionmentioning

confidence: 99%

“…Besides, conserved patterns and structure motifs can be found [1] by analyzing the distance or contact map. Predicted distance or contact maps are widely used in remote homology protein detection [2][3][4], protein information estimate [5,6], and protein structure research [7][8][9].…”

Section: Introductionmentioning

confidence: 99%

Freeprotmap: waiting-free prediction method for protein distance map

Huang,

Li,

Chen

et al. 2024

BMC Bioinformatics

View full text Add to dashboard Cite

Background Protein residue–residue distance maps are used for remote homology detection, protein information estimation, and protein structure research. However, existing prediction approaches are time-consuming, and hundreds of millions of proteins are discovered each year, necessitating the development of a rapid and reliable prediction method for protein residue–residue distances. Moreover, because many proteins lack known homologous sequences, a waiting-free and alignment-free deep learning method is needed. Result In this study, we propose a learning framework named FreeProtMap. In terms of protein representation processing, the proposed group pooling in FreeProtMap effectively mitigates issues arising from high-dimensional sparseness in protein representation. In terms of model structure, we have made several careful designs. Firstly, it is designed based on the locality of protein structures and triangular inequality distance constraints to improve prediction accuracy. Secondly, inference speed is improved by using additive attention and lightweight design. Besides, the generalization ability is improved by using bottlenecks and a neural network block named local microformer. As a result, FreeProtMap can predict protein residue–residue distances in tens of milliseconds and has higher precision than the best structure prediction method. Conclusion Several groups of comparative experiments and ablation experiments verify the effectiveness of the designs. The results demonstrate that FreeProtMap significantly outperforms other state-of-the-art methods in accurate protein residue–residue distance prediction, which is beneficial for lots of protein research works. It is worth mentioning that we could scan all proteins discovered each year based on FreeProtMap to find structurally similar proteins in a short time because the fact that the structure similarity calculation method based on distance maps is much less time-consuming than algorithms based on 3D structures.

show abstract

Reducing dimensionality in remote homology detection using predicted contact maps

Cited by 6 publications

References 31 publications

Remote-3DD: un nuevo método para la detección de homólogos remotos que usa propiedades fisicoquímicas

Remote-3DD: un nuevo método para la detección de homólogos remotos que usa propiedades fisicoquímicas

In silico analysis of phag-like protein in Ralstonia Euthropa H16, potentially involved in polyhydroxyalkanoates synthesis

Freeprotmap: waiting-free prediction method for protein distance map

Contact Info

Product

Resources

About