There are two tools that are proposed in the research.
- Prediction of spider neurotoxic peptide from amino acid sequences only.
- Data augmentation tool for peptide data
To install python dependencies for neurotoxic peptides, run: pip install -r requirements_neurotox_pred.txt
Input file type is fasta format in which amino acids are represented using single-letter codes.
Detailed information of fasta links: https://en.wikipedia.org/wiki/FASTA_format
Output file only contains peptide list which are predicted as neurotoxin.
python ntest.py --fasta <input_fasta_file.fasta> --output <output_file_name.csv>
To install python dependencies, run: pip install -r requirements.txt
Also, NCBI BLAST should be installed.
After install NCBI blast, BLAST bin path in aug_config.ini file should be modified.
Run
$ sudo apt-get install ncbi-blast
- Download latest BLAST version (Link)
- Decompress
$ tar -zxvf ncbi-blast-version_info-linux.tar.gz
Input file type is fasta format in which amino acids are represented using single-letter codes.
Detailed information of fasta links: https://en.wikipedia.org/wiki/FASTA_format
Output file contains information about probabilities of four respective ion-channel modulability.
Probability with >=0.5 predicts as modulator peptides for respective ion channels.
python peptide_augmentation.py --fasta <input_fasta_file.fasta> --cpu <max_cpu_usage> --eval <E-value cutoff>
Lee, Byungjo, Min K. Shin, In-Wook Hwang, Junghyun Jung, Yu J. Shim, Go W. Kim, Seung T. Kim, Wonhee Jang, and Jung-Suk Sung. 2021. "A Deep Learning Approach with Data Augmentation to Predict Novel Spider Neurotoxic Peptides" International Journal of Molecular Sciences 22, no. 22: 12291. https://doi.org/10.3390/ijms222212291