(1) 2020-08-27 valence parameter test re-fitting for fixing amide issue

1. new t70 child parameter candidate

`t68`(pre-existing)	`[:1]~[#7X3,#7X2-1:2]-[#6X3:3]~[:4]`
`t69`(pre-existing)	`[:1]~[#7X3,#7X2-1:2]-!@[#6X3:3]~[:4]`
`t70a` → `t69a`	`[:1]-[#7X3:2]-!@[#6X3$(=[#8,#16,#7]):3]~[*:4]` ( `-` or `-!@`?)
`t70`(pre-existing)	`[#1:1]-[#7X3:2]-[#6X3:3]=[#8,#16,#7:4]` ( `-` or `-!@`?)
`t70a`	`[*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#6,#1]`
`t70b`	`[#1:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#6,#1]`
`t70c`	`[*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#7X3]`

cf. Barrier height study (mainly for my own understanding) to understand (1) why we need child terms t70a and t70b .

- Conclusion: Based on the following plots, it seems reasonable to have separate child term [*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#1] (higher energy barrier) and [*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#7X3] ( lower energy barrier).

- No clear evidence for the necessity to combine a pattern with carbon attached to the carbonyl carbon ([*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#6]) and one with hydrogen attached to the carbonyl carbon([*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#1])

2. Test fitting

(1) re-fitting #1.1

Separated endo-cyclic and exo-cyclic rotations( t69a and t69b, t70 and t70a): will determine whether to separate or not after comparing the final k values.
w/ only selected td targets
w/ old conda env.(openforcefield==0.6.0, forcebalance=1.7.1)
optimized in 6 steps (3.21033e+01 → 7.97182e+00)

	SMIRKS	initial guess	fit #1.1

	SMIRKS	initial guess	fit #1.1
t69a	[:1]-[#7X3:2]-[#6X3$(=[#8,#16,#7]):3]~[*:4]	2.5 (1+cos(2x-180)) + 0.0 (1+cosx)	2.130536422094e+00 (1+cos(2x-180)) + 9.910621243621e-02 (1+cosx)
t69b	[:1]-[#7X3:2]-!@[#6X3$(=[#8,#16,#7]):3]~[*:4]	2.5 (1+cos(2x-180)) + 0.0 (1+cosx)	2.270075197860e+00 (1+cos(2x-180)) + 3.898905433309e-01 (1+cosx)
t70	[#1:1]-[#7X3:2]-[#6X3:3]=[#8,#16,#7:4]	3.459249459574e+00 (1+cos(2x-180)) + 1.356955617521e+00 (1+cosx)	3.471614493904e+00 (1+cos(2x-180)) + 1.353665035664e+00 (1+cosx)
t70a	[#1:1]-[#7X3:2]-!@[#6X3:3]=[#8,#16,#7:4]	2.5 (1+cos(2x-180)) + 2.0 (1+cosx)	2.5 (1+cos(2x-180)) + 2.0 (1+cosx)
t70b	[*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#6,#1]	2.5 (1+cos(2x-180)) + 0.0 (1+cosx)	2.779138641990e+00 (1+cos(2x-180)) + 3.381053683048e-02 (1+cosx)
t70c	[#1:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#6,#1]	2.5 (1+cos(2x-180)) + 2.0 (1+cosx)	2.590468938026e+00 (1+cos(2x-180)) + 1.820091344671e+00 (1+cosx)
t70d	[*:1]-[#7X3:2]-!@[#6X3:3](=[#8,#16,#7:4])-[#7X3]	2.5 (1+cos(2x-180)) + 0.0 (1+cosx)	1.888086753838e+0 (1+cos(2x-180)) - 2.028808021603e-01 (1+cosx)

QM vs MM torsion profiles