Table 4.

Affinities calculated by path sampling simulations.

	ΔG	σ	ε	${\mathrm{\Delta }\mathrm{G}}_{\mathrm{b}}^0$
WEHI-539
q1	16.16	0.01	0.02	−11.63
q2	16.89	0.01	0.02	−12.21
q3	22.44	0.01	0.02	−17.95
q4	20.26	0.01	0.02	−16.00
Exp	–	–	–	−12.66
ABT-737
q1	20.09	0.01	0.02	−15.45
q2	20.55	0.01	0.02	−16.12
q3	16.07	0.01	0.02	−11.56
q4	19.11	0.01	0.02	−14.50
Exp	–	–	–	−12.90

Affinities were taken from the 150 ns to 250 ns range (i.e. omitting the first 150 ns of simulation and only using the final 100 ns) for the WHAM analyses described in Tables S4–S11. The WHAM analyses were executed using the same parameters as the calculations used in the main text (i.e. a Δλ of 0.05 Å, a tolerance of 1e⁻⁸ and with 1000 bootstraps). ΔG is the average PMF over the final 50 bins (2.5 Å), with σ its standard deviation. ε corresponds to the average error (via bootstrapping) taken over the same range. ${\mathrm{\Delta }\mathrm{G}}_{\mathrm{b}}^0$ is the standard binding free energy calculated using ΔG and the corresponding sample COMs during the start and end range of the simulation following Eq. S7. The experimental (Exp) affinities (in the ${\mathrm{\Delta }\mathrm{G}}_{\mathrm{b}}^0$ column) were obtained from Lessene et al¹⁴.