Analytical PCR: generate expected PCR sizes algorithmically

[bthuronyi]

Using various strategies we can determine the expected size of an analytical PCR amplicon based on the parts list. Options for this:

1. Crude, but low user demand: assume mID-derived part lengths are equal to their donor plasmid's size minus a universal "plasmid backbone size" constant; assume dID-derived part lengths are equal to the dID size. Enforce that parts marked short are estimated at 0.1 kb if they're estimated to be longer. Assume primers are at the extreme end of each part and have length 20.
2. Better, moderate user demand: users select a Golden Gate donor backbone type when they register a construct, which corresponds to a Settings table of options -- or they input the actual backbone size, if not choosing an option. Use this to estimate part sizes. Use similar to above.
3. Pretty darn good, higher user demand: users paste the actual part sequence into the Registry when registering GG donors. Estimate band size by assuming primers are at extreme end of each part and have length 20.
4. Really good: above, but match the position of each primer to its part sequence and count those bases plus the ones between there and the part sequence end in the PCR product, plus the sequences of all intervening parts.

An advantage of 3-4 is that we could algorithmically generate Golden Gate plasmid sequences, which some users might greatly appreciate.

[bthuronyi]

Relatedly, options 3-4 would allow us to algorithmically suggest analytical PCR primers to go with GG parts in the Registry.