cpplocpred 1.0.0

Prediction of Cell-Penetrating Peptides and Subcellular Localization

CPPLocPred

Prediction of Cell-Penetrating Peptides and their Subcellular Localization

CPPLocPred is a two-stage machine learning tool:

• Stage 1 — CPP vs Non-CPP (ExtraTrees classifier, AAC features)
• Stage 2 — Subcellular localization (CatBoost classifiers, DDR features): Cytoplasm · Nucleus · Mitochondria · Endo/Lysosome · Others

Web server: https://webs.iiitd.edu.in/raghava/cpplocpred/ GitHub: https://github.com/namanm04/CPPLocPred

---

Installation

pip install cpplocpred

All model .pkl files, motif databases, and MERCI_motif_locator.pl are bundled inside the package — no separate download needed.

Dependencies installed automatically: pandas, scikit-learn, catboost

Motif search requires Perl to be installed and on your PATH.

---

Usage

Job 1 — CPP Prediction

cpplocpred -i input.fasta -o results.csv -j 1
cpplocpred -i input.fasta -o results.csv -j 1 -t 0.5

Job 2 — Motif Search

cpplocpred -i input.fasta -o motifs.csv -j 2 -l Nucleus
cpplocpred -i input.fasta -o motifs.csv -j 2 -l Mitochondria -c Koolman

All arguments

Flag	Description	Default
-i	Input FASTA file	required
-o	Output CSV file	required
-j	Job: 1=Prediction, 2=Motif Search	1
-t	CPP probability threshold	0.44
-m	Model directory	bundled in package
-l	Location: Cytoplasm, Nucleus, Mitochondria, Endo_lysosome, Others	Cytoplasm
-c	Motif class: None, Koolman, Betts-Russell, Rasmol	None
--motif_dir	Motif files directory	bundled in package
--perl	Path to perl executable	perl

---

Output

Job 1 — results.csv

Column	Description
ID	Sequence identifier
Sequence	Amino acid sequence
CPP_Probability	Stage 1 probability (0–1)
CPP_Prediction	CPP / Non-CPP / Invalid
Cytoplasm_Probability	Stage 2 localization score
Nucleus_Probability
Mitochondria_Probability
Endo_lysosome_Probability
Others_Probability
Final_Localization	Predicted location(s), ;-separated

Job 2 — motifs.csv + motifs_hits.csv

• motifs.csv — one row per sequence (ID, Length, Total_Hits, Motif_Patterns, …)
• motifs_hits.csv — one row per hit (ID, Start, End, Hit_Length, Motif_Pattern, Matched_Residues)

---

Input format

>seq1
RQIKIWFQNRRMKWKK
>seq2
GRKKRRQRRRPPQ

Only standard 20 amino acid single-letter codes accepted (ACDEFGHIKLMNPQRSTVWY).

---

Citation

Bajiya N., Mehta N.K., Raghava G.P.S. (2025) CPPLocPred: Machine learning-based prediction and subcellular localization of cell-penetrating peptides. IIIT Delhi. https://webs.iiitd.edu.in/raghava/cpplocpred/

---

License

MIT © 2025 Raghava Group, IIIT Delhi