Table 1|.
Summary of cryoEM datasets and hyperparameters used for classifier training on each. Each dataset was downsampled and split into train and test sets at the whole micrograph level.
| Train | Test | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Dataset | Protein | Original (ang/pix) | Down-sampled (ang/pix) | Particle radius (pix) | Training radius (pix) | π | Number of micro-graphs | Number of particles | Number of micro-graphs | Number of particles |
| EMPIAR-10025 | T20S proteasome | 0.98 | 15.7 | 7 | 3 | 0.035 | 156 | 39653 | 40 | 10301 |
| EMPIAR-10028 | 80S ribosome | 1.34 | 10.7 | 12 | 3 | 0.012 | 831 | 80701 | 250 | 24546 |
| EMPIAR-10096 | Hemagglu-tinin trimer | 1.31 | 5.24 | 10 | 4 | 0.035 | 347 | 100465 | 100 | 29535 |
| EMPIAR-10215 | Rabbit muscle aldolase | 0.832 | 6.64 | 10 | 3 | 0.1 | 865 | 163758 | 200 | 39347 |
| EMPIAR-10234 | Clustered protocad-herin | 1.061 | 8.49 | 15 | 4 | 0.015 | 67 | 1167 | 20 | 373 |
| Toll receptor | Toll receptor | 0.832 | 3.328 | 25 | 5 | 0.035 | 30 | 686 | 14 | 362 |