Server

A Web server has been set up to calculate the conservation of cysteines for a given protein sequence. Because of advances in sequence search methods the server applies a somewhat different approach to prepare the input sequence alignment than has been described in the original paper and used to calibrate the system (see Applied Databases and Methods). The submitted sequence is searched against a frequently updated non redundant (NR) database of sequences.37 The NR database currently contains more than 700,000 sequence entries. The sequence database search method uses PSI-BLAST38 with a strict (0.00001) e-value cutoff to extract significant hits. This default cutoff can be changed by the user if more remote homologs need to be included in the analysis and especially if no or only a few sequences have been found using the default values.

Once the PSI-BLAST search returns a set of homologous sequences and fragments, all the full-length sequences are extracted from the sequence database and aligned by CLUSTAL W.39 After the alignments are prepared the absolute and relative conservations of each position are calculated as described in Methods, using the approach of Livingstone and Barton as described in the AMAS program.25 The user can set the fraction of gaps that can be ignored at each position before the conservation scores are calculated. The Results page allows optional download of the full PSI-BLAST search. However, for practical reasons the output shows only a fraction of the used alignment, with the calculated absolute conservation values underneath. The absolute and relative conservation values for the cysteines are reported at the end of the Results page.

The Web server implementing the prediction method is accessible at http://guitar.rockefeller.edu/~andras/cysredox.html.

0 0

Post a comment