Zezima: Gene Prediction with Transformer Model

Overview

Zezima is a sophisticated gene prediction tool that leverages the power of Transformer models to analyze DNA sequences.
It's designed to efficiently predict gene-related features within DNA sequences, aiding in the complex task of genomic analysis.

Key Features

• Transformer Model: Utilizes an advanced machine learning approach for gene prediction.
• Custom DNA Sequence Handling: Tailored to process specific DNA sequence formats.
• Configurable Parameters: Offers flexibility to adjust model parameters in config.py

Requirements

Before beginning the setup, ensure your system meets the following requirements:

• Operating System: Linux
• Python Version: Python 3.10 or higher

Setting Up Python Virtual Environment

A Python virtual environment is recommended for managing the project's dependencies. Follow these steps to create and activate your virtual environment:

1. Create a Virtual Environment:

   python3.10 -m venv venv

2. Activate the Virtual Environment:

   source venv/bin/activate

3. Install Required Python Packages:

   pip3 install -Ur requirements.txt

Data Format

The input data file should be structured as follows, containing a header section and subsequent data vectors:

#HEADER#
#DATE=2024-02-25T12:02:29.238927
#pre_processing_version=[0, 1, 0]
#bp_vector_schema=['A', 'C', 'G', 'T', 'PROMOTOR_MOTIF', 'ORF', 'POLY_ADENYL', 'miRNA', 'rRNA', 'gene']
#description of nucleotide:A=[1, 0, 0, 0], C=[0, 1, 0, 0], G=[0, 0, 1, 0], T=[0, 0, 0, 1]
#description of feature:0=no_present, 1=start, 2=continuation/ongoing, 3=end
#max_feature_overlap=0
####END####
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0]	[0, 1, 0, 0, 0, 1, 0, 0, 0, 0]

Each vector in the file represents information about a specific position on the DNA, as defined by bp_vector_schema. The vector encodes the presence and status of various genetic features at that position.

For more detailed information about the input data structure and the pre-processing steps, please refer to the pre-processing documentation.

Usage

1. Prepare your DNA sequence data files according to the specified format and place them in the designated input directory. In a case that you don't have your own pipelines you can use or modify this pre-processing stage to your needs

2. Configure the model and execution parameters in the config.py file.

3. Run the application:

   python3.10 run.py

License

Zezima is open-source software licensed under the MIT License. For more details, see the LICENSE file.