.
** Server did not say "OK".
Your options are to try again later, use the Blast email server
(add '-mail' to the command line), or use
Blast via the web.
The old 'blast -batch' command doesn't work anymore; to run
blast in batch mode see
How do I run blast
searches in batch mode
Why have my alignments changed in Version 9?
The following information is taken from the
GCG Newsletter, July 1997.
Protein sequence alignment programs in the Wisconsin Package use scoring
matrices to make comparisons between pairs of amino acids. In Version 9.0,
released in December 1996, changes were made to the scoring matrices which
may affect the alignments you create. If the sequences you aligned in
previous versions of the Package were very similar, you may not have noticed
much difference to those same alignments produced using Version 9.0.
However, with more distantly related sequences the alignments produced may
be substantially different.
Version 9 Changes to Scoring Matrices
There are two main changes to protein scoring matrices in Version 9.0.
- BLOSUM621, the new default protein scoring matrix, is the primary cause
for the difference in alignments created in Version 9.0. Versions of the
Wisconsin Package prior to 9.0 used a scoring matrix based on the PAM250
table from M. Dayhoff2, where all perfect matches were set to a single
value. BLOSUM62, in contrast, assigns scores ranging from 4 to 11 to perfect
matches (identity). Scores for a non-perfect but similar matches range from 0
to 3. Scores
for mismatches (non-similar matches) are assigned vales less than 0.
Knowing the range of these
values is helpful for setting display thresholds for similar amino acids.
- Scoring matrix values are now integers. Matrices which previously held
floating point (real number) values have had their values multiplied by 10
so that they now hold integer values. In addition, the values for gap
creation and extension penalties also must be entered as integer values.
The modified PAM250 matrix which was used in pr