Prediction of LL-37 localization in its precursor protein: l is the rightmost a.a. in the sliding window, pAntimicrobial is probability of antimicrobial activity, L' is the sliding window length (A – 20, B – 37 and C – 50 amino acid a.a.)

Machine learning using multiple logistic regression for antimicrobial and hemolytic peptides prediction and their identification in large proteins

Figure 1 - Prediction of LL-37 localization in its precursor protein:

l is the rightmost a.a. in the sliding window, pAntimicrobial is probability of antimicrobial activity, L' is the sliding window length (A – 20, B – 37 and C – 50 amino acid a.a.)

Prediction of LL-37 localization in its precursor protein: l is the rightmost a.a. in the sliding window, pAntimicrobial is probability of antimicrobial activity, L' is the sliding window length (A – 20, B – 37 and C – 50 amino acid a.a.)

the actual mature LL-37 length is 37 a.a. The dashed horizontal line represents the estimated optimal cut-off for AMPs prediction (0.6)

DOI: 10.60797/jbg.2024.26.5.12