Antimicrobial Peptide

From the long and short independent test sets, we extracted all PeptideDB peptide sequences. We refer to this set as the PeptideDB.70 dataset, as no peptide sequence in this dataset has 70% sequence similarity to any peptide in the training set. We divided these peptides into three activity subsets i.e. antimicrobial peptides, peptide hormones and toxin/venom peptides (see Table S12). We used the three activity subsets to compare the predictive power of PeptideRanker and two other state-of-the-art antimicrobial peptide predictors in each of these three bioactive peptide classes.

The benchmark dataset $\mathbb{S}$ used in this study can be formulated as
$\mathbb{S}={\mathbb{S}}^{+}\cup {\mathbb{S}}^{-}$
where the positive and negative subsets, ${\mathbb{S}}^{+}$ and ${\mathbb{S}}^{-}$ , contain respectively anticancer and non-anticancer peptides, while the symbol ∪ represents the union in the set theory. As elucidated by a comprehensive review [31 (link)], there is no need to separate the benchmark dataset into a training dataset and a testing dataset if the predictor to be developed will be tested by the jackknife test or subsampling (K-fold) cross-validation test since the outcome thus obtained is actually from a combination of many different independent dataset tests. In order to have a high quality benchmark dataset, the samples in the positive subset were taken from Hajisharifi et al. [16 (link)] that contain 138 anticancer peptides, which had been derived from the antimicrobial peptide database [62 (link)] as well as the existing literatures. The samples in the negative subset, however, were constructed as follows. In view of the fact that the peptides with anticancer activity are generally secretory [63 (link)], the non-anticancer peptides can be selected from the non-secretory proteins deposited in Universal Protein Resource. To avoid redundancy and reduce the homology bias, peptides with more than 90% sequence similarity were removed by using the CD-HIT program [64 (link)]. After such a screening procedure, we finally obtained 206 non-anticancer peptides for the negative subset. The 138 anticancer peptides and 206 non-anticancer peptides are given in Supporting Information S1.
The statistical distribution of the length for the 138 anticancer peptides is given in Figure 3, from which we can see that most (∼80%) of them are with the length less than 30 amino acids.
As clearly pointed out in the beginning of this section, the independent dataset is not absolutely needed for validating a predictor via the jackknife or K-fold cross-validation, but as a demonstration to show how to use the proposed predictor, it may be of help [65 (link)] to also construct an independent dataset ${\mathbb{S}}_{\text{Ind}}$ as formulated by
${\mathbb{S}}_{\text{Ind}}={\mathbb{S}}_{\text{Ind}}^{+}\cup {\mathbb{S}}_{\text{Ind}}^{-}$
where the samples in ${\mathbb{S}}_{\text{Ind}}^{+}$ and ${\mathbb{S}}_{\text{Ind}}^{-}$ were fetched from the dataset used by Tyagi et al. [15 (link)] and the recent CancerPPD database [66 (link)] according to the following criteria: (1) none of the anticancer peptides in ${\mathbb{S}}_{\text{Ind}}^{+}$ occurs in ${\mathbb{S}}^{+}$ ; (2) none of the non-anticancer peptides in ${\mathbb{S}}_{\text{Ind}}^{-}$ occurs in ${\mathbb{S}}^{-}$ ; (3) neither the included peptides in ${\mathbb{S}}_{\text{Ind}}$ contains illegal single-letter amino acid codes such as “B”, “U”, “X”, and “Z”, nor the peptides in ${\mathbb{S}}_{\text{Ind}}$ has ≥ 90% pairwise sequence identity to any other in the benchmark dataset of Eq.1. By strictly following the aforementioned procedures, we finally obtained an independent dataset ${\mathbb{S}}_{\text{Ind}}$ , in which the positive subset ${\mathbb{S}}_{\text{Ind}}^{+}$ contains 150 anticancer peptides, and the negative subset ${\mathbb{S}}_{\text{Ind}}^{-}$ contains 150 non-anticancer peptides. See Supporting Information S2 for the detailed information. Actually, all the datasets used in this study can also be directly downloaded from the website at http://lin.uestc.edu.cn/server/iACP/data.

Antimicrobial peptide susceptibility Human neutrophil defensin-1 (hNP-1) (AnaSpec Incorporated, California, USA) and LL-37 (Sigma) susceptibility assays were performed as described previously^{55 (link)}. Briefly, overnight cultures were normalised to an OD_600nm of 0.1 and incubated with 5 μg/mL of hNP-1 or LL-37 for 2 h at 37 °C. Serial dilutions were plated on tryptic soy agar (TSA) to determine CFUs. The same number of bacterial cells inoculated into PBS, diluted, and plated acted as a control. Survival was determined as the percentage of CFU on exposure to either LL-37 or HNP-1 relative to the PBS control. Relative survival was determined through normalisation to JE2.

Candidate genes associated with immunity, reproduction, and growth were screened using a keyword search of the BLASTx-annotated PANM-DB. The names of representative genes involved in various stages of insect immunity, cell signaling, sex-determination, reproduction and growth processes were included as keywords. Further, the GO terms and KEGG classifications were also referred to identify putative functional transcripts. A comprehensive network of immunity-related transcripts were screened and was categorized into ‘Pathogen Recognition Receptor (Immune signaling pathway)’, ‘TLR Signaling Pathway (Adapter proteins, MyD88-dependent pathway)’, ‘Endogenous Ligands’, ‘Immune Effectors’, ‘Antimicrobial Peptides’, ‘Cytokines and Cytokine Receptors’, ‘Apoptosis’ and ‘Autophagy’.