TABLE 2.

Distance-restrained docking: application to RNAP

	Experimental FRET data			Simulated FRET data
	Rif	Rif SV	Rif+Rif SV	Rif	Rif SV	Rif+Rif SV
Number of restraints	21	21	42	21	21	42
Accuracy (Å)	7.1	10.1	7.1	11±4	13 ± 3	8 ± 2
Precision (Å)	7.4	9.3	6.6	12 ± 7	12 ± 4	9 ± 4
Mean FRET penalty (Å)	0.9	1.1	1.0	0.4 ± 0.3	0.2 ± 0.1	0.4 ± 0.2
% FRET violation (0–5 Å)	48	57	36	65 ± 14	80 ± 7	72 ± 8
% FRET violation (5–10 Å)	9	10	17	24 ± 8	16 ± 7	18 ± 7
% FRET violation (>10 Å)	43	33	47	11 ± 12	4 ± 5	10 ± 5

Comparison of model quality for models of antibiotics generated with experimental and simulated FRET data. Simulated FRET efficiency data were generated by back-calculating exact donor-acceptor distances in the RNAP holoenzyme reference model and incorporating 10% systematic error as well as 15% random error into the restraints. Results are postprocessed to eliminate all sterically impossible solutions given the RNAP reference model.