Table 4.

Summary of the feedback loop statistics for the Tc holotoxin data.

Feedback round	Good classes	Good particles	Picks/Mic	Good picks/Mic	Resolution	Relative good picks
1	130	12,595	130	25	4.28	0.19
2	145	10,406	100	21	4.24	0.21
3	151	14,534	74	29	4.36	0.40
4	146	14,081	71	28	4.28	0.40
5	155	14,935	68	27	4.28	0.40
6	140	13,566	69	27	4.24	0.39
6 + T 0.194	145	13,954	55	28	4.24	0.50

For every feedback round as well as the final run after optimization of the picking threshold (6 + T x.xx) the number of classes labeled “good” by Cinderella; the number of particles included in these classes; the total number and the number of good particles picked per micrograph; the final resolution of the 3D reconstruction; and the relative amount of good particles (highlighted in bold) are listed for 500 micrographs.