Extracting physicochemical features to predict protein secondary structure.

We propose a protein secondary structure prediction method based on position-specific scoring matrix (PSSM) profiles and four physicochemical features including conformation parameters, net charges, hydrophobic, and side chain mass. First, the SVM with the optimal window size and the optimal parameters of the kernel function is found. Then, we train the SVM using the PSSM profiles generated from PSI-BLAST and the physicochemical features extracted from the CB513 data set. Finally, we use the filter to refine the predicted results from the trained SVM. For all the performance measures of our method, Q 3 reaches 79.52, SOV94 reaches 86.10, and SOV99 reaches 74.60; all the measures are higher than those of the SVMpsi method and the SVMfreq method. This validates that considering these physicochemical features in predicting protein secondary structure would exhibit better performances.

Discovering weighted patterns in intron sequences using self-adaptive harmony search and back-propagation algorithms.

A hybrid self-adaptive harmony search and back-propagation mining system was proposed to discover weighted patterns in human intron sequences. By testing the weights under a lazy nearest neighbor classifier, the numerical results revealed the significance of these weighted patterns. Comparing these weighted patterns with the popular intron consensus model, it is clear that the discovered weighted patterns make originally the ambiguous 5SS and 3SS header patterns more specific and concrete.

Modelling splice sites with locality-sensitive sequence features.

The splice sites are essential for pre-mRNA maturation and crucial for Splice Site Modelling (SSM); however, there are gaps between the splicing signals and the computationally identified sequence features. In this paper, the Locality Sensitive Features (LSFs) are proposed to reduce the gaps by homogenising their contexts. Under the skewness-kurtosis based statistics and data analysis, SSM attributed with LSFs is fulfilled by double-boundary outlier filters. The LSF-based SSM had been applied to six model organisms of diverse species; by the accuracy and Receiver Operating Characteristic (ROC) analysis, the promising results show the proposed methodology is versatile and robust for the splice-site classification. It is prospective the LSF-based SSM can serve as a new infrastructure for developing effective splice-site prediction methods and have the potential to be applied to other sequence prediction problems.

Intron identification approaches based on weighted features and fuzzy decision trees.

Current computational predictions of splice sites largely depend on the sequence patterns of known intronic sequence features (ISFs) described in the classical intron definition model (IDM). The computation-oriented IDM (CO-IDM) clearly provides more specific and concrete information for describing intron flanks of splice sites (IFSSs). In the paper, we proposed a novel approach of fuzzy decision trees (FDTs) which utilize (1) weighted ISFs of twelve uni-frame patterns (UFPs) and forty-five multi-frame patterns (MFPs) and (2) gain ratios to improve the performances in identifying an intron. First, we fuzzified extracted features from genomic sequences using membership functions with an unsupervised self-organizing map (SOM) technique. Then, we brought in different viewpoints of globally weighting and crossly referring in generating fuzzy rules, which are interpretable and useful for biologists to verify whether a sequence is an intron or not. Finally, the experimental results revealed the effectiveness of the proposed method in improving the identification accuracy. Besides, we also implemented an on-line intronic identifier to infer an unknown genomic sequence.

Finding new core promoter elements using backward-looking strategies.

Core Promoter Elements (CPEs) were key players in transcription initiation. Identifying CPEs is crucial for understanding gene expression. In this paper, a framework for finding new CPEs was proposed. An experiment was performed on the sequences of Eukaryotic Promoter Database (EPD). From the results, the known CPEs were all recovered; in addition, five new motifs were discovered in Drosophila and three in human. By comparing the results with currently known CPEs, it is shown that the proposed system is feasible and reliable, and these new CPEs are worth of further exploration.