I want to call the attention of non-virologists interested
in protein sequence alignments and their significance to a paper
of mine that just appeared in JOURNAL OF GENERAL VIROLOGY (VOL. 71,
p. 1009). The paper reports the successful alignment of very distantly
related amino acid sequences of proteins putatively involved in
the cell-to-cell transport of infections in plants. I am interested
in reactions to two non-standard approaches used in that paper.
1. The first is the alignment method. It is not completely
objective but produces alignments that on first inspection seem
superior to both CLUSTAL and MULTALIN produced alignments. I
would appreciate general comments on the validity of the method.
2. The second non-standard approach is the determination
of significance scores by comparison of the test alignment with
alignments with one sequence off-set by one or more positions
relative to the other sequence. The method does not destroy
inherent periodicities in the occurrence of specific residue types.
The significance scores appear to correlate better with PAM scores
than do significance scores calculated by the use of scrambled
amino acid sequences. Has this territory already been explored?
Should I further explore the utility of this measure of significance?
Relative to both approaches, I would like to know whether
there is any interest in having the programs developed for general
use (the programs currently run slowly on an Apple II and have
several errors). Is anyone interested in helping in such develop-
ment?
Thanks in advance for any input you may care to give.
Ulrich Melcher
BIOCUKM at OSUCC
