Table 1:

Prokaryotic benchmark datasets statistics. Code column indicates the subcellular location representation in our predictive model. Gram-negative bacteria have five major subcellular localization sites, namely, the cytoplasm, the periplasm, the inner membrane, the outer membrane, and the extracellular space, whereas Gram-positive bacteria do not have an outer cell membrane. However in these benchmark datasets cell wall is absent in Gram-negative dataset and in Gram-positive bacteria, we observe the lack of periplasm proteins.

No	Subcellular location	Code	Proteins count
			Gram negative	Gram positive
1	Cytoplasm	C	4,152	349
2	Extracellular	S	272	290
3	Inner membrane	I	1,415	1,779
4	Outer membrane	O	346	–
5	Periplasm	P	422	–
6	Cell wall	W	–	34
7	Vacuole	V	10	4
Multiple localizations			39	8
Total			6,578	2,448