Tests multiple sequence alignment methods. PREFAB is a test set which exploits methodology, test data and statistical methods that have previously been applied to alignment accuracy assessment. It contains (i) a set of reference pair alignments, (ii) samples of sequences for testing multiple alignment programs and (iii) a program of assessing the quality of work of multiple alignment programs.
Provides high quality reference alignments based on 3D structural superpositions. BAliBASE is a large scale benchmarks specifically designed for multiple sequence alignment. The alignment test cases are manually refined to ensure the correct alignment of conserved residues. The alignments are organized into reference sets designed to represent real multiple alignment problems.
Allows study of both sequence and structure relationships between homologous proteins. HOMSTRAD is a database that provides combined protein sequence and structure information extracted from the Protein Data Bank (PDB), a primary protein structure repository, and relies heavily on other databases, especially Pfam and SCOP. It contains about 2700 families, and just under half of which are multi-member.
Provides a data resource of structural alignments. SAD chooses a good alignment from several candidates via an alignment quality measure. This database uses alignments from multiple other databases and an ab initio method named internal coordinate mechanics (ICM) global structural alignment. It uses also the contact alignment quality score, a scoring function, to assess trial alignments by their statistical significance associated with their ability to reproduce conserved three-dimensional residue contacts.