Protein Sequence Descriptor Converter

This Python script converts protein sequences into numerical descriptors based on amino acid properties. It supports two descriptor types:

1. VHSE (Vector of Hydrophobic, Steric, and Electronic properties): Based on the paper: https://doi.org/10.1002/bip.20296
2. Z-scales: Based on the paper: https://pubs.acs.org/doi/full/10.1021/jm9700575

Requirements

• Python 3.x
• pandas
• scikit-learn

To install the required packages:

pip install pandas scikit-learn

Files

• descriptor_converter.py: Main Python script
• example.csv: Input file containing protein sequences (one sequence per line, no header)
• data_vhse.csv: Output file with calculated descriptors

Usage

1. Prepare your input file (example.csv) with one protein sequence per line
2. Run the script:

python descriptor_converter.py

3. The script will generate data_vhse.csv containing the numerical descriptors

Customization

You can modify the descriptor type by changing the parameter in the script:

# For VHSE descriptors (8 values per amino acid)
data_descriptors = get_descriptors(data_list, descriptor='vhse')

# For Z-scales descriptors (5 values per amino acid)
data_descriptors = get_descriptors(data_list, descriptor='zscales')

Functions

• assign_descriptor(sequence, descriptor): Converts a single protein sequence to descriptors
• get_descriptors(data_list, descriptor): Processes a list of sequences
• norm_features(data): Normalizes the descriptors using MinMaxScaler

Output

The output file contains numerical descriptors where:

• Each row represents one protein sequence
• Each column represents a descriptor value
• For VHSE: 8 values per amino acid
• For Z-scales: 5 values per amino acid

Example

Input protein sequence: "ACDE"

• With VHSE, this would generate 4×8 = 32 descriptor values
• With Z-scales, this would generate 4×5 = 20 descriptor values

Citation

If using this tool for research, please cite the relevant descriptor papers:

• VHSE: Mei H, Liao ZH, Zhou Y, Li SZ. A new set of amino acid descriptors and its application in peptide QSARs. Biopolymers. 2005.
• Z-scales: Sandberg M, Eriksson L, Jonsson J, Sjöström M, Wold S. New chemical descriptors relevant for the design of biologically active peptides. A multivariate characterization of 87 amino acids. Journal of Medicinal Chemistry. 1998.