Computational prediction and analysis of protein γ-carboxylation sites based on a random forest method.

The glutamate γ-carboxylation plays a pivotal part in a number of important human diseases. However, traditional protein γ-carboxylation site detection by experimental approaches are often laborious and time-consuming. In this study, we initiated an attempt for the computational prediction of protein γ-carboxylation sites. We developed a new method for predicting the γ-carboxylation sites based on a Random Forest method. As a result, 90.44% accuracy and 0.7739 MCC value were obtained for the training dataset, and 89.83% accuracy and 0.7448 MCC value for the testing dataset. Our method considered several features including sequence conservation, residual disorder, secondary structures, solvent accessibility, physicochemical/biochemical properties and amino acid occurrence frequencies. By means of the feature selection algorithm, an optimal set of 327 features were selected; these features were considered as the ones that contributed significantly to the prediction of protein γ-carboxylation sites. Analysis of the optimal feature set indicated several important factors in determining the γ-carboxylation and a possible consensus sequence of the γ-carboxylation recognition site (γ-CRS) was suggested. These may shed some light on the in-depth understanding of the mechanisms of γ-carboxylation, providing guidelines for experimental validation.

[Processing and analysis of ChIP-seq data].

The next-generation sequencing coupled with chromatin immunoprecipitation (ChIP-seq) is becoming a key technology for the study of transcriptional regulation in the context of functional genomics. Due to the overwhelming amount of data generated from ChIP-seq experiments, the ChIP-seq data processing brings many new challenges in the field of bioinformatics. Considering the development of data processing skills largely behind that of the ChIP-seq experiment techniques, it is urgent to give a review on the ChIP-seq data processing for more and more oncoming researchers to build or improve algorithms. This paper provides a brief overview of the ChIP-seq data processing, highlighting the main prob-lems and methods in detail, to allow scientists to understand rapidly and deeply.

Studies on the rules of ß-strand alignment in a protein ß-sheet structure.

To further disclose the underlying mechanisms of protein ß-sheet formation, studies were made on the rules of ß-strands alignment forming ß-sheet structure using statistical and machine learning approaches. Firstly, statistical analysis was performed on the sum of ß-strands between each ß-strand pairs in protein sequences. The results showed a propensity of near-neighbor pairing (or called "first come first pair") in the ß-strand pairs. Secondly, based on the same dataset, the pairwise cross-combinations of real ß-strand pairs and four pseudo-ß-strand contained pairs were classified by support vector machine (SVM). A novel feature extracting approach was designed for classification using the average amino acid pairing encoding matrix (APEM). Analytical results of the classification indicated that a segment of ß-strand had the ability to distinguish ß-strands from segments of α-helix and coil. However, the result also showed that a ß-strand was not strongly conserved to choose its real partner from all the alternative ß-strand partners, which was corresponding with the ordination results of the statistical analysis each other. Thus, the rules of "first come first pair" propensity and the non-conservative ability to choose real partner, were possible important factors affecting the ß-strands alignment forming ß-sheet structures.

Prediction of function changes associated with single-point protein mutations using support vector machines (SVMs).

Computational methods can be used to predict the effects of single amino acid substitutions (single-point mutations). In contrast to previous methods that need many protein sequence and structural features, we applied support vector machines (SVMs) to predict protein function changes associated with amino acid substitutions using only sequence information, and cross-validated them on a large dataset extracted from the Protein Mutant Database (PMD). By three SVM classifiers, we investigated three local sequence features of proteins (residue composition, hydrophobic interaction, and evolutionary property), and examined their effects on the prediction accuracy. As a main result, a novel SVM named substitution-matrix-based kernel SVM was constructed to make speedy and accurate prediction, and its value was shown in an application case. Furthermore, our findings confirmed results from other studies.