Table 2.

Overlap measures for the eight perturbations of NOA and OA, performed at the MM level, based on 60,000 (NOA) or 4000 (OA) snapshots. Each measure is the minimum (Ω, K _AB, and $\Pi$) or maximum (w _max, ΔΔG _EA, and ΔΔG _TI) value over the 26 λ values for the simulations with and without the host.

	Ω	K AB	Π	w max	ΔΔG EA	ΔΔG TI
NOA
MeBz→Bz	1.00	1.03	2.9	0.003	0.02	0.04
EtBz→MeBz	1.00	1.04	2.7	0.004	0.03	0.03
pClBz→Bz	1.00	1.03	2.9	0.002	0.02	0.04
mClBz→Bz	1.00	1.03	2.8	0.002	0.06	0.02
Hx→Bz	1.00	1.03	2.5	0.004	0.06	0.02
MeHx→Hx	1.00	1.03	2.8	0.017	0.05	0.01
Hx→Pen	1.00	1.04	2.5	0.026	0.08	0.04
Hep→Hx	1.00	1.04	2.5	0.009	0.05	0.06
OA
MeBz→Bz	0.99	1.01	2.3	0.010	0.20	0.13
EtBz→MeBz	1.00	1.02	2.1	0.024	0.23	0.03
pClBz→Bz	1.00	1.02	2.5	0.009	0.12	0.14
mClBz→Bz	0.99	1.02	2.2	0.019	0.24	0.02
Hx→Bz	0.99	1.00	1.2	0.106	0.94	0.21
MeHx→Hx	0.93	0.79	1.1	0.947	16.19	0.34
Hx→Pen	0.95	0.85	0.0	0.716	1.69	0.06
Hep→Hx	0.98	0.93	−0.1	0.377	6.42	0.03

The measures are the Bhattacharyya coefficient for the energy distribution overlap (Ω),73 the Wu & Kofke overlap measures of the energy probability distributions (K _AB) and their bias metrics (Π)74, 75 the weight of the maximum term in the EA (w _max),20 the difference of the forward and backward EA estimate (ΔΔG _EA in kJ/mol), and the difference between the BAR and TI estimates (ΔΔG _TI in kJ/mol). Values indicating poor overlap or bad convergence are marked in bold face (Ω < 0.7, K _AB < 0.7, $\Pi$ < 0.5, w _max > 0.2, ΔΔG _EA > 4 kJ/mol, or ΔΔG _TI > 1 kJ/mol).72, 74, 75