NeuroPIpred is a machine learning-based computational platform developed for predicting, designing, and scanning insect neuropeptides.
The platform uses experimentally validated natural and modified insect neuropeptides for building predictive models.
NeuroPIpred can:
- Predict insect neuropeptides
- Design mutant neuropeptide analogs
- Scan proteins for neuropeptide regions
- Perform similarity search using BLAST
Title: NeuroPIpred: a tool to predict, design and scan insect neuropeptides
Authors:
Piyush Agrawal, Sumit Kumar, Archana Singh, Gajendra P. S. Raghava, and Indrakant K. Singh
Journal: Scientific Reports (2019)
Insect neuropeptides are signaling molecules that regulate:
- Mating
- Migration
- Oviposition
- Metabolism
- Growth and development
- Homeostasis
These peptides are promising targets for insect pest control and insecticide development.
Two datasets were developed:
- 875 unique natural insect neuropeptides
- Random SwissProt peptides as negative dataset
- 2024 modified neuropeptides with C-terminal amidation
- Modified peptides from SATPDB as negative dataset
The following techniques were used:
- Support Vector Machine (SVM)
- Random Forest
- SMO
- J48 Decision Tree
- Naive Bayes
Residue frequency-based features.
Captures residue composition and local residue order.
Encodes residue order and positional information.
Uses N-terminal and C-terminal residues separately.
Dipeptide composition based SVM model:
- Validation Accuracy: 83.71%
- MCC: 0.67
Dipeptide composition based SVM model:
- Validation Accuracy: 97.93%
- MCC: 0.96
Residues enriched:
- C
- D
- F
- G
- N
- S
- Y
Residues enriched:
- D
- E
- F
- G
- M
- N
- P
- R
- S
- Y
Important motifs identified:
- ECC
- QCK
- FDEI
- EIDR
- GPR
- SFGL
- WFGP
- YSF
Predicts neuropeptides from peptide sequences.
Generates mutant analogs for activity optimization.
Identifies possible neuropeptide regions in proteins.
Finds similar experimentally validated neuropeptides.
Allows downloading benchmark datasets.
- SVM_light
- WEKA
- Random Forest
- MERCI
- PSI-BLAST
NeuroPIpred can be used for:
- Insect neuropeptide discovery
- Pest control research
- Peptide engineering
- Bioactive peptide design
- Insecticide development
| Method | Accuracy | MCC |
|---|---|---|
| NeuroPID | 52.57% | 0.16 |
| NeuroPIpred | 83.71% | 0.67 |
Web Server:
https://webs.iiitd.edu.in/raghava/neuropipred/
Docker Image:
raghavagps/gpsrdocker
Department of Computational Biology
Indraprastha Institute of Information Technology Delhi
New Delhi, India
Email: [email protected]
Creative Commons Attribution License
Generated from the uploaded NeuroPIpred research paper.