Detección de homólogos remotos usando propiedades fisicoquímicas

Bedoya, Oscar

doi:10.24050/reia.v14i27.1161

Cited by 2 publications

(3 citation statements)

References 17 publications

(18 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The classifiers considered in this research use 3D models enriched with physicochemical properties. Every protein is represented as the number of times that each model is observed in a predicted contact map and an interaction matrix as presented in Bedoya & Tischer (2015). Submitting a test sample to the whole set of binary classifiers produces a score-vector.…”

Section: The Single-mcs Methodsmentioning

confidence: 99%

“…Table 1 also shows the specific ROC score reached for each superfamily when different methods for detecting remote homologs are used. In the remote-3DI method (Bedoya & Tischer, 2015) every protein is represented using a predicted contact map and an interaction matrix. The remote-3DI method uses models with 3D information (i.e., typical 3D interactions that occur in the contact map) enriched with physicochemical properties.…”

Section: The Hierarchical-mcs Methodsmentioning

confidence: 99%

“…We select the classifier with the highest ROC score using 5-fold cross validations on the training data set. We chose between the same classification techniques (BayesNet, NaiveBayes, NaiveBayes Multinomial, Multilayer perceptron, Hyper pipes, VFI, LMT) that were considered for detecting remote homologs in Bedoya and Tischer (2015). Then, a multi-class SCOP class predictor is obtained by using the same steps that we used in the single-MCS method but considering SCOP classes instead of SCOP superfamilies.…”

Section: Evaluating the Hierarchical-mcs Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Predicción de superfamilias usando modelos 3D enriquecidos con propiedades fisicoquímicas

Bedoya

Tischer

2016

IyC

View full text Add to dashboard Cite

In this paper, two new methods that address the multi-class superfamily prediction problem are presented. In the multi-class superfamily recognition problem each amino acid sequence has to be classified into one of the known structural classes (i.e., superfamilies). Most of the strategies that have been proposed to predict superfamilies are based on using the binary classifiers that detect remote homologs. The remote homology detection problem is about finding a classifier that is able to separate remote homologs from non-remote homologs. The current methods for multi-class superfamily recognition take the outputs of the binary classifier (i.e., the scores) for each SCOP superfamily in the data set and build a classification model (i.e., multi-class classifier). Unlike the current methods, which represent a protein considering the amino acids composition, in this research we use the number of times that 3D models enriched with physicochemical properties occur in both its predicted contact map and its interaction matrix. We hypothesize that including both 3D information and physicochemical properties might have an impact in the accuracy obtained during the superfamily prediction. In this paper, we present two new strategies for predicting superfamilies that use 3D models enriched with physicochemical properties, the single-MCS and the hierarchical-MCS methods, which reach an accuracy percentage of 74% and 76% on the SCOP 1.53 data set, respectively. In addition, tests on the SCOP 1.55 and the SCOP 1.61 are also presented. Keywords: 3D enrich models, Binary classifiers, Physicochemical properties, SCOP superfamily, Superfamily prediction. ResumenEn este artículo se presenta dos nuevos métodos para la predicción de superfamilias. En el problema de la predicción de superfamilias cada secuencia de aminoácidos se debe clasificar en una de las clases estructurales conocidas (i.e., superfamilias). La mayoría de las estrategias que se han propuesto para predecir superfamilias se basan en usar los clasificadores binarios que detectan homólogos remotos. Detectar homólogos remotos está relacionado con encontrar un clasificador que es capaz de indicar si una proteína es, o no, un homólogo remoto de un conjuntos de proteínas. Los métodos actuales para detectar superfamilias toman las salidas de los clasificadores binarios para cada superfamilia y construyen un modelo de clasificación. A diferencia de los métodos actuales, los cuales representan a las proteínas considerando la composición de aminoácidos, nosotros usamos el número de veces que modelos 3D enriquecidos con propiedades fisicoquímicas ocurren tanto en el mapa de contacto predicho como en la matriz de interacción. Nuestra hipótesis es que al incluir los modelos 3D con las propiedades fisicoquímicas se puede tener un impacto en la exactitud obtenida durante la predicción de superfamilias. En este artículo se presenta dos nuevas estrategias para predecir superfamilias, los métodos single-MCS y hierarchical-MCS, los cuales alcanzan una exactitud del 74% y 76% e...

show abstract

Section: The Single-mcs Methodsmentioning

confidence: 99%

Section: The Hierarchical-mcs Methodsmentioning

confidence: 99%

Section: Evaluating the Hierarchical-mcs Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Predicción de superfamilias usando modelos 3D enriquecidos con propiedades fisicoquímicas

Bedoya

Tischer

2016

IyC

View full text Add to dashboard Cite

show abstract

PrESOgenesis: A two-layer multi-label predictor for identifying fertility-related proteins using support vector machine and pseudo amino acid composition approach

Bakhtiarizadeh

Rahimi

Mohammadi‐Sangcheshmeh

et al. 2018

Sci Rep

View full text Add to dashboard Cite

Successful spermatogenesis and oogenesis are the two genetically independent processes preceding embryo development. To date, several fertility-related proteins have been described in mammalian species. Nevertheless, further studies are required to discover more proteins associated with the development of germ cells and embryogenesis in order to shed more light on the processes. This work builds on our previous software (OOgenesis_Pred), mainly focusing on algorithms beyond what was previously done, in particular new fertility-related proteins and their classes (embryogenesis, spermatogenesis and oogenesis) based on the support vector machine according to the concept of Chou’s pseudo-amino acid composition features. The results of five-fold cross validation, as well as the independent test demonstrated that this method is capable of predicting the fertility-related proteins and their classes with accuracy of more than 80%. Moreover, by using feature selection methods, important properties of fertility-related proteins were identified that allowed for their accurate classification. Based on the proposed method, a two-layer classifier software, named as “PrESOgenesis” (https://github.com/mrb20045/PrESOgenesis) was developed. The tool identified a query sequence (protein or transcript) as fertility or non-fertility-related protein at the first layer and then classified the predicted fertility-related protein into different classes of embryogenesis, spermatogenesis or oogenesis at the second layer.

show abstract

Detección de homólogos remotos usando propiedades fisicoquímicas

Cited by 2 publications

References 17 publications

Predicción de superfamilias usando modelos 3D enriquecidos con propiedades fisicoquímicas

Predicción de superfamilias usando modelos 3D enriquecidos con propiedades fisicoquímicas

PrESOgenesis: A two-layer multi-label predictor for identifying fertility-related proteins using support vector machine and pseudo amino acid composition approach

Contact Info

Product

Resources

About