The production of a trait may be many different genes (not sure if multiple produce thioalchol). Also, Staphylococcus hominis is a prokaryotic cell, and therefore the procedure would be much different than if you wanted eukaroyote-eukaryote. There is going to be a much different procedure each time, but most of the time you can check out KEGG for pathways. Perhaps, for example, take a look at
http://www.genome.jp/kegg-bin/show_organism?org=sab to get BLAST comparisons. Also, I'm pretty sure that they've figured out the traits that produce thioalcohol (
https://www.sciencedaily.com/releases/2015/03/150330213947.htm , wikipedia link). You want to replace thioalcohol with ester, so naturally you want insertion to be over the thioalcohol. Alternatively, you can just replicate as a plasmid, albeit with less stability. Would allow you to get rid of the 'homology' parts of design below.
Your description of replacing the gene that produces thioalcohol with one that produces ester is not a project of identifying and isolating genes but unfortunately one of cloning, transforming, and troubleshooting. Not quite as exciting. No identification is needed (since they've found the genes) and isolation is a vague description (deleting, transcriptionally isolating, isolating from genome, orthogonal isolation, or basic PCR?). Isolating a gene for production would just be PCR and is pretty easy and robust, you can buy stuff off of eBay to do that (although I'd recommend a better polymerase than taq)
I know that Staphylococcus aureus CRISPR is characterized, so what I would do is target thioalcohol to be cut with the CRISPR from Staphylococcus (in which you would supply the Staphylococcus CRISPR on a plasmid)
... [homology] [thioalcohol] [homology]...
\ \ cut^ cut^ \ \
\ \ \ \
-Plasmid-( [homology] [ester] [homology] ) - [sgRNA] - [CRISPR locus] -
The sgRNA, homology, and ester would likely have to by synthesized. That's the easiest way to do that, in my opinion. Sorry if the drawing is bad.
-Koeng