Welcome to TandemMod’s documentation!

TandemMod is a deep learning framework for detecting RNA modifications in nanopore direct RNA sequencing (DRS) data. It offers three modes: de novo training, transfer learning, and prediction. Researchers can train from scratch, fine-tune pre-trained models, or apply existing models for predictions. TandemMod achieves high accuracy and can be used to identify multi tpyes of RNA modifications in various species and conditions. It provides a user-friendly solution for studying RNA modifications.

To use TandemMod, you can follow the instructions below:

Contents

• Installation
• Data preprocessing
  • Basecalling
  • Multi-fast5 to single-fast5
  • Resquiggling
  • Feature extraction
• Usage
  • Source code
  • Training from scratch
  • Transfer learning
  • Prediction
• Run examples
  • Train m6A model using IVET m6A dataset
  • Train m6A model using curlcake m6A dataset
  • Transfer m6A model to m7G using ELIGOS dataset
  • Predict m6A sites in human cell lines
• Data postprocessing
  • Transcriptome location to genome location
  • Read-level predictions to site-level predictions
  • Futher analysis
• Getting Help
• Citing TandemMod

Citing TandemMod

If you use TandemMod in your research, please cite **********************************

Contacts

TandemMod is developed and maintained by You Wu and Xiang Yu from Shanghai Jiao Tong University. If you want to contribute or have any questions, please leave an issue in our repository.

Thank you!