The benchmark dataset contains 266 mycobacterial non-secretory protein sequences and 35 mycobacterial secretory protein sequences. These protein sequences share the sequence identity of <30%.

Download