The Toolkit for Quick Reproducibility & Utility

An interactive guide to IonNTxPred. Move beyond static documentation and learn how to predict, analyze, and engineer ion channel-impairing proteins with ease.

What is IonNTxPred?

This section introduces the core purpose and capabilities of IonNTxPred. It's designed to give you a quick yet comprehensive overview of what you can achieve with the tool, from basic classification to advanced protein engineering, helping you understand which feature best suits your research needs.

🔮 Prediction

Classify proteins as ion channel-impairing or non-impairing based on their primary sequence.

🔬 Protein Scanning

Pinpoint specific "toxic" regions or hotspots within a protein that are responsible for its activity.

🛠️ Protein Design

Computationally generate and evaluate all possible single-point mutants to engineer novel proteins.

🧩 Motif Scanning

Utilize the MERCI algorithm to identify known functional motifs related to toxicity within your sequence.

A high-level overview of the IonNTxPred workflow.

Installation Guide

Getting IonNTxPred running on your system is the first step. This section provides clear, actionable instructions for two installation methods. Follow the recommended 'Pip' method for a quick setup, or use 'Conda' for a more isolated and reproducible environment.

Option 1: Pip (Recommended)

The quickest way to get started. Installs the package and its dependencies from the Python Package Index.

pip install ionntxpred

ionntxpred -h

Option 2: Standalone with Conda

For a fully controlled environment, clone the repository and use the provided environment file.

1. Clone the repository:

git clone https://github.com/raghavagps/IonNTxPred.git
cd IonNTxPred

2. Create & activate environment:

conda env create -f environment.yml
conda activate IonNTxPred

⚠️ Critical Step: Download The Models

The trained models are too large for GitHub or PyPI. You **must** download them separately. The tool will not work without them.

Visit the Official Download Page.
Download the `model.zip` file.
**Unzip the file** inside your main `IonNTxPred` directory before running any analysis.

Interactive Command Guide

This is the heart of the interactive guide. Instead of just reading about command-line options, you can build your exact command here. Select your desired job, choose a model, and set your parameters. The tool will generate a ready-to-use command, eliminating guesswork and typos.

1. Select Job Type

2. Select Model

3. Select Target Ion Channel

Your Generated Command:

ionntxpred.py -i input.fasta -o output.csv

Remember to replace `input.fasta` and `output.csv` with your actual file names.

Input & Output

✅ Input File Format

IonNTxPred accepts two simple formats:

FASTA Format: The standard for sequences, with a `>` header line followed by the sequence.
Simple Format: A plain text file with one protein sequence on each new line.

✅ Output File

Results are always saved in a clean, universal **CSV (Comma-Separated Values)** format.

This makes it easy to open your results in spreadsheet software like Excel or to parse them for further analysis with scripts in Python or R.

🔍 Recommended thresholds

Based on our analysis, we suggest using the following optimized thresholds for best performance on each ion channel type:

Dataset	Classifier	Recommended Threshold
Na⁺	ESM2-t33	0.21
Na⁺	Hybrid Model (ESM2-t33 + BLAST)	0.41
K⁺	ESM2-t33	0.21
K⁺	Hybrid Model (ESM2-t33 + BLAST)	0.21
Ca²⁺	ESM2-t33	0.35
Ca²⁺	Hybrid Model (ESM2-t33 + BLAST)	0.35
Other	ESM2-t33	0.30
Other	Hybrid Model (ESM2-t33 + BLAST)	0.30

📌 Note: These thresholds were determined through rigorous testing on independent datasets and are recommended for a balanced sensitivity and specificity.

This page serves as a comprehensive resource for users of IonNTxPred, offering valuable assistance in effectively utilizing its diverse prediction modules. Within this guide, you will find detailed information, figures, and step-by-step instructions on leveraging each module integrated into IonNTxPred. We aim to equip you with a clear understanding of how to navigate and make the most of every module within the platform.