FScanpy

A Machine Learning-Based Framework for Programmed Ribosomal Frameshifting Prediction

FScanpy is a comprehensive Python package designed for the prediction of Programmed Ribosomal Frameshifting (PRF) sites in nucleotide sequences. By integrating advanced machine learning approaches (Gradient Boosting and BiLSTM-CNN) with the established FScanR framework, FScanpy provides robust and accurate PRF site predictions. The package requires input sequences to be in the positive (5' to 3') orientation.

For detailed documentation and usage examples, please refer to our tutorial.

Core Features

• Sequence Feature Extraction: Support for extracting features from nucleic acid sequences, including base composition, k - mer features, and positional features.
• Frameshift Hotspot Region Prediction: Predict potential PRF sites in nucleotide sequences using machine learning models.
• Feature Extraction: Extract relevant features from sequences to assist in prediction.
• Cross - Species Support: Built - in databases for viruses, marine phages, Euplotes, etc., enabling PRF prediction across various species.

Main Advantages

• High Accuracy: Integrates multiple machine learning models to provide accurate PRF site predictions.
• Efficiency: Utilizes a sliding window approach and feature extraction techniques to rapidly scan sequences.
• Versatility: Supports PRF prediction across various species and can be combined with the FScanR framework for enhanced accuracy.
• User - Friendly: Comes with detailed documentation and usage examples, making it easy for researchers to use.
• Flexible: Provides different resolutions to suit different using situations.

Installation Requirements

• Python ≥ 3.7
• Dependencies are automatically handled during installation

Option 1: Install via pip

pip install FScanpy

Option 2: Install from source

git clone git@60.204.158.188:yyh/FScanpy-package.git
cd FScanpy-package
pip install -e .

Authors

Citation

If you utilize FScanpy in your research, please cite our work:

[Citation details will be added upon publication]