🧬 PyRNA-Lab

AI-Powered RNA Secondary Structure Prediction Toolkit

PyRNA-Lab is an open-source Python toolkit for RNA sequence analysis and AI-assisted RNA secondary structure prediction.

It combines Python, PyTorch, computational biology, and an interactive Pygame interface into a simple research and educational environment for experimenting with RNA sequences.

---

✨ Features

• 🧬 RNA sequence input and analysis
• 🤖 AI-powered RNA secondary structure prediction
• 🧠 PyTorch neural-network model
• 🔗 RNA base-pair prediction
• 📐 Secondary structure representation
• 📊 Sequence length statistics
• 🔬 Selected base-pair scores
• 🖥️ Interactive graphical interface
• 🌙 Dark / Light theme
• 📋 Copyable model output
• 🖱️ Text selection with mouse
• ⌨️ Keyboard shortcuts
• 📜 Scrollable sequence and output areas
• 📦 Python package available through PyPI
• ⚡ pyrna command-line launcher
• 🔓 Open-source project

---

🚀 Installation

PyRNA-Lab can be installed directly from PyPI.

Requirements

• Python 3.10 or newer
• PyTorch
• NumPy
• Pandas

Install the package:

pip install pyrna-lab

For AI dependencies:

pip install pyrna-lab[ai]

---

🖥️ Launching PyRNA-Lab

After installation, simply run:

pyrna

The PyRNA-Lab graphical interface will open automatically.

You do not need to navigate into the project directory or manually execute app.py.

---

🧬 Using PyRNA-Lab

1. Enter an RNA sequence

When the application starts, the RNA input area is displayed on the left side.

You can:

• Type an RNA sequence manually
• Paste a sequence from the clipboard
• Select text with the mouse
• Move through the sequence with the keyboard
• Delete characters
• Select all text
• Copy selected text
• Cut selected text
• Scroll horizontally and vertically

The input field does not automatically modify your text.

PyRNA-Lab does not automatically:

• Convert lowercase to uppercase
• Remove spaces
• Remove new lines
• Filter characters
• Truncate the input

The original user input remains unchanged.

---

🧪 Supported RNA Bases

The current AI prediction model accepts the standard RNA bases:

A
C
G
U

Examples:

GGGAAAUCCCGGAUUU

AUGCUAGCUAGCUA

GGCAUCGGAUCC

If an unsupported character is present, PyRNA-Lab reports an error instead of silently modifying the sequence.

For example:

Invalid RNA base: X

---

🤖 Running an AI Prediction

After entering an RNA sequence, click:

Predict Structure

PyRNA-Lab then:

1. Reads the provided RNA sequence.
2. Encodes RNA bases numerically.
3. Sends the encoded sequence to the PyTorch model.
4. Evaluates potential nucleotide pairs.
5. Applies valid RNA pairing rules.
6. Builds a secondary structure representation.
7. Displays the predicted structure.
8. Displays selected base pairs and their scores.
9. Generates a complete copyable output.

---

🔗 RNA Base-Pair Rules

PyRNA-Lab currently considers the following base pairs:

A - U
U - A

C - G
G - C

G - U
U - G

These correspond to standard Watson-Crick pairings plus the GU wobble pairing used by the current prediction workflow.

---

🧠 AI Model

PyRNA-Lab loads its trained neural-network model from:

rna_pair_model.pth

The model is loaded automatically when the application starts.

The interface displays the model status:

● AI READY

when the model loads successfully.

If the model cannot be loaded, the interface displays:

● MODEL ERROR

and prediction is disabled.

---

📏 Current Model Limitation

The current trained model supports a maximum of:

128 bases

This is a model limitation, not an input-field limitation.

The application itself does not restrict the amount of text that can be entered.

However, if the sequence is longer than 128 bases, PyRNA-Lab will not run the prediction and will display:

The current AI model supports up to 128 bases

---

📐 Predicted Structure

After a successful prediction, PyRNA-Lab displays the generated RNA secondary structure in the:

PREDICTED STRUCTURE

panel.

The structure uses bracket notation such as:

((...))

and may contain additional bracket types depending on the generated structure.

The interface visually distinguishes structural brackets and unpaired positions.

---

📊 Prediction Statistics

After prediction, the application displays three statistics.

RNA Length

Shows the number of bases in the input sequence.

Example:

RNA LENGTH
16

Base Pairs

Shows the number of selected base pairs identified by the prediction workflow.

Example:

BASE PAIRS
5

Model

Displays:

AI RNA

to indicate that the RNA AI model is being used.

---

📋 Model Output

The right side of the interface contains a complete model-output area.

After prediction, it contains:

RNA SEQUENCE
============

[RNA sequence]


PREDICTED STRUCTURE
===================

[Predicted structure]


SELECTED BASE PAIRS
===================

Position     Score
------------------------
  1 -  16      0.842
  2 -  15      0.731

The exact results depend on the input sequence and model prediction.

---

🔗 Selected Base Pairs

Each selected pair contains:

• First nucleotide position
• Second nucleotide position
• Model score

For example:

Position     Score
------------------------
  1 -  16      0.842
  2 -  15      0.731

Positions are displayed using 1-based indexing, meaning the first nucleotide is position 1.

---

📋 Copying Results

The model output is read-only.

You cannot modify the generated model output.

However, you can select and copy it.

Mouse

Drag across the output text to select it.

Select everything

Press:

Ctrl + A

Copy

Press:

Ctrl + C

The selected output can then be pasted into another application.

---

⌨️ Keyboard Shortcuts

RNA Input

Shortcut	Action
Ctrl + A	Select all input
Ctrl + C	Copy selected input
Ctrl + X	Cut selected input
Ctrl + V	Paste
Ctrl + Enter	Run prediction
Escape	Clear input and results
Left Arrow	Move cursor left
Right Arrow	Move cursor right
Home	Move to beginning
End	Move to end
Up / Down	Scroll
Page Up	Scroll up
Page Down	Scroll down
Backspace	Delete previous character
Delete	Delete next character

---

Model Output

Shortcut	Action
Ctrl + A	Select all output
Ctrl + C	Copy selected output
Mouse drag	Select output
Mouse wheel	Scroll output

The output area does not allow typing, pasting, cutting, deleting, or modifying the model-generated result.

---

🧹 Clearing the Workspace

Click:

Clear

or press:

Escape

while working inside the RNA input.

This resets:

• RNA sequence
• Prediction output
• Predicted structure
• Base-pair count
• RNA length
• Previous sequence
• Scroll position
• Status message

The application returns to:

Ready

---

🌙 Dark and Light Themes

PyRNA-Lab includes two visual themes.

Dark

The default theme provides a dark interface with a gold visual accent.

Light

Click the theme button in the upper-right corner to switch to the light interface.

The button changes between:

● Bright

and:

● Dark

depending on the current theme.

The theme can be switched at any time.

---

🖱️ Scrolling

Scrollable areas are available for longer content.

RNA Input

The sequence editor supports:

• Vertical scrolling
• Horizontal scrolling
• Mouse wheel
• Shift + mouse wheel

Model Output

The output area supports:

• Vertical scrolling
• Horizontal scrolling
• Mouse wheel
• Shift + mouse wheel

This allows longer sequences and prediction outputs to be inspected without modifying the generated data.

---

🔬 Example Workflow

A typical workflow is:

1. Launch PyRNA-Lab
        │
        ▼
2. Enter RNA sequence
        │
        ▼
3. Click "Predict Structure"
        │
        ▼
4. Sequence encoding
        │
        ▼
5. PyTorch AI model
        │
        ▼
6. Base-pair evaluation
        │
        ▼
7. Structure generation
        │
        ▼
8. View predicted structure
        │
        ▼
9. Inspect base pairs & scores
        │
        ▼
10. Copy results if needed

---

🧪 Example

Start PyRNA-Lab:

pyrna

Enter:

GGGAAAUCCCGGAUUU

Then click:

Predict Structure

PyRNA-Lab will process the sequence and display:

• RNA length
• Selected base pairs
• Pair scores
• Predicted secondary structure
• Complete copyable output

---

🐍 Python Package Usage

PyRNA-Lab can also be imported as a Python package.

import pyrna_lab as rna

print(rna)

The package can therefore be used as part of a Python environment in addition to its graphical application.

---

💻 Running from Source

Clone the repository:

git clone https://github.com/paknahad91/PyRNA-Lab.git

Enter the project:

cd PyRNA-Lab

Install it in editable mode:

pip install -e .

Launch the application:

pyrna

---

📁 Project Structure

PyRNA-Lab/
│
├── rna_lab/
│   ├── app.py
│   ├── pair_model.py
│   ├── structure.py
│   │
│   ├── rna_model.pth
│   ├── rna_pair_model.pth
│   └── baseline_paired_unpaired.pth
│
├── README.md
├── pyproject.toml
└── ...

---

🛠️ Technology Stack

Technology	Role
Python	Core programming language
PyTorch	AI / neural-network inference
Pygame	Graphical user interface
NumPy	Scientific computing
Pandas	Data processing
PyPI	Python package distribution

---

🧬 Architecture Overview

                 PyRNA-Lab
                     │
          ┌──────────┴──────────┐
          │                     │
    Graphical UI           Python Package
          │                     │
       Pygame              pyrna_lab
          │
          ▼
    RNA Sequence
          │
          ▼
    Sequence Encoding
          │
          ▼
     PyTorch Model
          │
          ▼
    Pair Evaluation
          │
          ▼
   Structure Generation
          │
       ┌──┴───┐
       ▼      ▼
 Structure   Base Pairs
       │      │
       └──┬───┘
          ▼
      Model Output

---

⚠️ Error Messages

Please enter an RNA sequence

The input field is empty.

Enter an RNA sequence before running prediction.

---

Invalid RNA base: X

The sequence contains a character that the current model does not support.

Use:

A
C
G
U

---

The current AI model supports up to 128 bases

The sequence contains more than 128 positions.

Shorten the sequence or use a future model version that supports longer sequences.

---

AI model is not ready

The trained model could not be loaded.

Check that the model file is included correctly in the installation.

---

Prediction failed

An unexpected error occurred while running the prediction pipeline.

Check the terminal output for additional technical information.

---

🔐 Input & Output Behavior

PyRNA-Lab intentionally keeps the user's input unchanged.

The application does not silently:

• Uppercase the sequence
• Remove spaces
• Remove line breaks
• Filter characters
• Modify the original sequence

Validation occurs when prediction is requested.

The generated model output is read-only and can only be selected and copied.

---

🎯 Project Vision

PyRNA-Lab aims to provide an accessible computational environment for experimentation with:

• RNA sequence analysis
• RNA secondary structure prediction
• Machine learning
• Deep learning
• Computational biology
• Scientific programming
• AI-assisted biological research

The project is designed to make experimentation with RNA and AI approachable for students, developers, researchers, and computational biology enthusiasts.

---

🔬 Research Direction

Future research and development may explore:

• Improved RNA structure prediction
• Larger training datasets
• More advanced neural architectures
• Model benchmarking
• Better sequence representations
• Uncertainty estimation
• Explainable AI
• Longer sequence support
• Additional RNA analysis tools

These are research directions and should not be interpreted as features currently implemented in version 1.0.0.

---

📌 Project Status

Version: 1.0.0

Development Status: Alpha / Research Project

PyRNA-Lab is an evolving research-oriented project. APIs, models, and internal implementation details may change in future releases.

---

📦 Package Information

Package name:

pyrna-lab

Python import:

import pyrna_lab

Command-line launcher:

pyrna

---

🌐 Links

GitHub

https://github.com/paknahad91/PyRNA-Lab

PyPI

https://pypi.org/project/pyrna-lab/

---

👨‍💻 Author

Mohammad Paknahad

PyRNA-Lab is an independent project focused on combining:

Python + AI + RNA + Computational Biology

---

📄 License

PyRNA-Lab is released under the MIT License.

---

⭐ Support PyRNA-Lab

If you find PyRNA-Lab useful:

• ⭐ Star the GitHub repository
• 🐛 Report bugs
• 💡 Suggest improvements
• 🔬 Experiment with the toolkit
• 🧬 Use it for educational and research purposes

---