You could align the sequences for the virulence factor(s) of interest and construct a maximum likelihood phylogenetic tree. If you have metadata (phenotypic data etc) for the organisms from which the sequences were obtained, you could overlay these on the tree and see if you can observe some patterns such as clade associated with certain phenotypes etc. You could then use this information to identify the specific base changes in the sequence alignment that are associated with the observed pattern in the tree.

There are many db containing proteins involved in virulence: How To Find Proteins Related To Patogenicity Of A Bacteria ? or also PHI-base.

However, I will focus on which factor you want consider. This might help you to make a more focus and precise project. You can start reading some papers and reviews on this topic.