[PFScience]

Hello all.....a query from a relative newb!

When doing a retro-analysis is there an approximate rule of thumb as to when you should be looking to perform FGI's?

Are they something that you should be looking to do nearer the start of the process or nearer the end?

I know that a C=O or C-X or C-OH are good to aim for as these themselves can be converted into most of the other FG's. So realistically does that mean that you should only have a couple of FGI's in any given process?....Accepting obviously that the larger a molecule gets, the more complex retro-analysis can be.

Thanks

[phth]

IMO one should think about the bond disconnections theoretically, and figure out a couple pathways that may work.  Furthermore, assignment of the forward synthetic pathway with reagents and FGI's is when a plan can become infeasible with certain or all FGI's.  The more FGI's in one step the better the synthesis is (e.g. manipulation of alkenes by oxidative cleavage can install multiple functional groups in 1 step...), or the less number of columns (practically).