In the past, nearest neighbor algorithms for learning from examples have worked best in domains in which all features had numeric values. In such domains, the examples can be treated as points and distance metrics can use standard definitions. In symbolic domains, a more sophisticated treatment of the feature space is required. We introduce a nearest neighbor algorithm for learning in domains with symbolic features. Our algorithm calculates distance tables that allow it to produce real-valued distances between instances, and attaches weights to the instances to further modify the structure of feature space. We show that this technique produces excellent classification accuracy on three problems that have been studied by machine learning researchers: predicting protein secondary structure, identifying DNA promoter sequences, and pronouncing English text. Direct experimental comparisons with the other learning algorithms show that our nearest neighbor algorithm is comparable or superior ...

We describe improvements of the currently most popular method for prediction of classically secreted proteins, SignalP. SignalP consists of two different predictors based on neural network and hidden Markov model algorithms, where both components have been updated. Motivated by the idea that the cleavage site position and the amino acid composition of the signal peptide are correlated, new features have been included as input to the neural network. This addition, combined with a thorough error-correction of a new data set, have improved the performance of the predictor significantly over SignalP version 2. In version 3, correctness of the cleavage site predictions have increased notably for all three organism groups, eukaryotes, Gram-negative and Grampositive bacteria. The accuracy of cleavage site prediction has increased in the range from 6-17 % over the previous version, whereas the signal peptide discrimination improvement is mainly due to the elimination of false positive predictions, as well as the introduction of a new discrimination score for the neural network. The new method has also been benchmarked against other available methods. Predictions can be made at the publicly available web server

Artificial neural networks have been applied to the prediction of splice site location in human pre--mRNA. A joint prediction scheme where prediction of transition regions between introns and exons regulates a cutoff level for splice site assignment was able to predict splice site locations with confidence levels far better than previously reported in the literature. The problem of predicting donor and acceptor sites in human genes is hampered by the presence of numerous amounts of false positives --- in the paper the distribution of these false splice sites is examined and linked to a possible scenario for the splicing mechanism in vivo. When the presented method detects 95% of the true donor and acceptor sites it makes less than 0.1% false donor site assignments and less than 0.4% false acceptor site assignments. For the large data set used in this study this means that on the average there are one and a half false donor sites per true donor site and six false acceptor ...

Translation in eukaryotes does not always start at the first AUG in an mRNA, implying that context information also plays a role. This makes prediction of translation initiation sites a non-trivial task, especially when analysing EST and genome data where the entire mature mRNA sequence is not known. In this paper, we employ artificial neural networks to predict which AUG triplet in an mRNA sequence is the start codon. The trained networks correctly classified 88 % of Arabidopsis and 85 % of vertebrate AUG triplets. We find that our trained neural networks use a combination of local start eodon context and global sequence information. Furthermore, analysis of false predictions shows that AUGs in frame with the actual start codon are more frequently selected than out-of-frame AUGs, suggesting that our networks use reading frame detection. A number of conflicts between neural network predictions and database annotations are analysed in detail, leading to identification of possible database errors.

A hidden Markov model of signal peptides has been developed. It contains submodels for the N-terminal part, the hydrophobic region, and the region around the cleavage site. For known signal peptides, the model can be used to assign objective boundaries between these three regions. Applied to our data, the length distributions for the three regions are significantly different from expectations. For instance, the assigned hydrophobic region is between 8 and 12 residues long in almost all eukaryotic signal peptides. This analysis also makes obvious the difference between eukaryotes, Gram-positive bacteria, and Gram-negative bacteria. The model can be used to predict the location of the cleavage site, which it finds correctly in nearly 70% of signal peptides in a cross-validated test---almost the same accuracy as the best previous method. One of the problems for existing prediction methods is the poor discrimination between signal peptides and uncleaved signal anchors, but this is substant...

We describe a new classifier for protein secondary structure prediction which is formed by cascading together different types of classifiers using neural networks and linear discrimination. The new classifier achieves an accuracy of 76.7% (assessed by a rigorous full Jack-knife procedure) on a new non-redundant dataset of 496 non-homologous sequences (obtained from G.J. Barton and J.A. Cuff). This database was especially designed to train and test protein secondary structure prediction methods, and it uses a more stringent definition of homologous sequence than in previous studies. We show that it is possible to design classifiers which can highly discriminate the 3 classes (H, E, C) with an accuracy of up to 78% for b-strands, using only a local window and resampling techniques. This indicates that the importance of long range interactions for the prediction of b-strands has been previously overestimated. 2 Introduction Although the protein folding process may require catalysts suc...

This study describes a new Hidden Markov Model (HMM) system for segmenting uncharacterized genomic DNA sequences into exons, introns, and intergenic regions. Separate HMM modules were designed and trained for specific regions of DNA: exons, introns, intergenic regions, and splice sites. The models were then tied together to form a biologically feasible topology. The integrated HMM was trained further on a set of eukaryotic DNA sequences, and tested by using it to segment a separate set of sequences. The resulting HMM system, which is called VEIL (Viterbi Exon-Intron Locator), obtains an overall accuracy on test data of 92% of total bases correctly labelled, with a correlation coefficient of 0.73. Using the more stringent test of exact exon prediction, VEIL correctly located both ends of 53% of the coding exons, and 49% of the exons it predicts are exactly correct. These results compare favorably to the best previous results for gene structure prediction, and demonstrate the benefits of...

this paper we address the organism-specific aspects of the problem and present neural-network based prediction methods to identify signal peptides and their cleavage sites in protein sequences from Gram-positive and Gramnegative bacteria, humans and other eukaryotes.

Genes in eukaryotic DNA cover hundreds or thousands of base pairs, while the regions of those genes that code for proteins may occupy only a small percentage of the sequence. Identifying the coding regions is of vital importance in understanding these genes. Many recent research efforts have studied computational methods for distinguishing between coding and noncoding regions, and several promising results have been reported. We describe here a new approach, using a machine learning system that builds decision trees from the data. This approach combines several coding measures to produce classifiers with consistently higher accuracies than previous methods, on DNA sequences ranging from 54 base pairs to 162 base pairs in length. The algorithm is very efficient, and it can easily be adapted to different sequence lengths. Our conclusion is that decision trees are a highly effective tool for identifying protein coding regions. Keywords: coding regions, decision trees, machine learning, ex...

When preparing data sets of amino acid or nucleotide sequences it is necessary to exclude redundant or homologous sequences in order to avoid overestimating the predictive performance of an algorithm. For some time methods for doing this have been available in the area of protein structure prediction. We have developed a similar procedure based on pair-wise alignments for sequences with functional sites. We show how a correlation coefficient between sequence similarity and functional homology can be used to compare the efficiency of different similarity measures and choose a non-arbitrary threshold value for excluding redundant sequences. The impact of the choice of scoring matrix used in the alignments is examined. We demonstrate that the parameter determining the quality of the correlation is the relative entropy of the matrix, rather than the assumed (PAM or identity) substitution model. Results are presented for the case of prediction of cleavage sites in signal peptides. By inspec...