Table 1. Summary of the atomic composition of all the protein fragments used for EEM model calibration.

System	RS1	RS2	T1	T2	T3	T4	T5
Atoms	23259	42295	1167	1125	1065	1040	1075
Fragments	57	43	1	1	1	1	1
C	6649	12522	350	342	305	298	260
H	10434	20995	562	505	498	503	558
N	2730	3339	117	129	123	104	75
O	3332	5225	136	144	136	133	176
S	30	149	0	3	2	0	6
Ca	84	65	2	2	1	2	0

Reference sets RS1 and RS2 were used for model calibration, and internal and external validation. Five additional test molecules T1–T5 were used for external validation.