Background: The most common substitution matrices currently used (BLOSUM and PAM) are based on protein sequences with average amino acid distributions, thus they do not represent ...
Background: As in many different areas of science and technology, most important problems in bioinformatics rely on the proper development and assessment of binary classifiers. A ...
Background: There are many important clustering questions in computational biology for which no satisfactory method exists. Automated clustering algorithms, when applied to large,...
Background: Interest is growing in the application of syntactic parsers to natural language processing problems in biology, but assessing their performance is difficult because di...
Background: There is a large amount of microarray data accumulating in public databases, providing various data waiting to be analyzed jointly. Powerful kernel-based methods are c...