What exactly are you having trouble with? ClustalW is a fantastic program, but most biologists that use it from time to time don't bother to look up the help given that most of the parameters are optimized for the average user's needs. The program does, however, provide a number of parameters to optimize for the specific protein/gene that you are trying to align. For example, if you are doing an intraspecific (within a species) comparison, you might want to give more weight to a neutral nucleotide substitution (one that doesn't change the amino acid sequence) than if you were doing an interspecific comparison. My advice would be not to worry about the myriad options that ClustalW provides - you're not publishing your 'results' and you're likely to mess things up if you change one of the parameters without knowing a WHOLE lot about the comparison you're trying to make.
And the PIM:
"percentage identity" = some self-important person's jargon for percentage similarity
"matrix" = the % is based on the matrix of parameters that you have provided: gap length penalty, etc.
Oh, and you might want to actually check the results of the alignment - see if the program has gotten the sequences to line up right or if it has inserted some gaps for no good reason. As fantastically complex as this thing is, it sometimes (read, often) makes mistakes. It's still a vast improvement over trying to do it yourself - the Western mind was not meant to look at a sequence of ACTG and see anything other than CAT or TAG or CATTAG.
™ Version 4.2
