In Silico Identification of Antimicrobial Peptide Candidates in the Camellia sinensis Proteome

Creative Commons License

Eminoğlu A.

12. Uluslararası Ege Sağlık ve Fen Bilimleri Kongresi, İzmir, Türkiye, 25 - 26 Şubat 2026, ss.1, (Özet Bildiri)

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: İzmir
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.1
  • Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
  • Recep Tayyip Erdoğan Üniversitesi Adresli: Evet

Özet

ABSTRACT

Plant antimicrobial peptides (AMPs) are generally cysteine-rich peptides with molecular weights of 2–10 kDa and high structural stability. Although Camellia sinensis exhibits strong antimicrobial activity, this activity is largely attributed to phenolic compounds, and structurally defined bona fide AMPs from this species have not yet been explicitly reported. This study was designed as a preliminary analysis covering the first 500 proteins of the tea proteome to evaluate the applicability of a screening approach developed to identify AMP candidates. Protein sequences were imported into the R environment in FASTA, segmented into 20–40 amino acid subsequences using a sliding window approach, and evaluated using support vector machine (SVM), Random Forest (RF), artificial neural network (ANN), and discriminant analysis (DA) based classifiers implemented in the CAMP database. A total of 3,035 high-confidence candidates predicted as AMPs by all four classifiers were selected. For each candidate, amino acid composition-based metrics were calculated, including net-charge, isoelectric point (pI), hydrophobicity (GRAVY), fractions of positively charged and hydrophobic residues, amphipathicity, and alpha-helix propensity. Initial filtering was performed using rule-based criteria frequently reported for AMPs in the literature (net charge +2 to +5, pI ≥ 8, GRAVY −1 to +1, amphipathic-proxy ≥ 0.4, and minimum cysteine content), followed by multi-criteria prioritization based on length, charge, and composition. The high-confidence AMP candidates exhibited positive net charge (median = 3, range: 2–5), basic pI (median ≈ 8.8), balanced hydrophobicity (median GRAVY ≈ 0.35), pronounced amphipathicity (median ≈ 0.51), and high alpha-helix propensity (median ≈ 0.96). Classifier probability scores further indicated high prediction confidence (SVM median = 0.81, RF = 0.73, DA = 0.91).These findings demonstrate that peptides derived from the tea proteome can be identified as high-confidence AMP candidates using in silico approaches and that these candidates are consistent with the characteristic biophysical profile of α-helical AMPs.

Keywords: Antimicrobial Peptides, Camellia sinensis, Biophysical Characterization.