| # TReconLM | |
| TReconLM is a decoder-only transformer model for trace reconstruction of noisy DNA sequences. It is trained to reconstruct a ground-truth sequence from multiple noisy copies (traces), each independently corrupted by insertions, deletions, and substitutions. | |
| ## Model Variants | |
| We provide pretrained and fine-tuned model checkpoints for the following ground-truth sequence lengths: | |
| - L = 60 | |
| - L = 110 | |
| - L = 180 | |
| Each model supports reconstruction from cluster sizes between 2 and 10. | |
| ## How to Use | |
| A Colab notebook is available in our [GitHub repository](https://github.com/MLI-lab/TReconLM) under `trace_reconstruction.ipynb`, which demonstrates how to load the model and run inference on our benchmark datasets. The test datasets used in the notebook can be downloaded from [Hugging Face](https://huggingface.co/datasets/mli-lab/TReconLM_datasets). | |
| ## Training Details | |
| - Models are pretrained on synthetic data generated by sampling ground-truth sequences of length L uniformly at random over the quaternary alphabet, and independently introducing insertions, deletions, and substitutions at each position. | |
| - Error probabilities for insertions, deletions, and substitutions are drawn uniformly from the interval [0.01, 0.1], and cluster sizes are sampled uniformly from [2, 10]. | |
| - Models are fine-tuned on real-world sequencing data (Noisy-DNA and Microsoft datasets). | |
| For full experimental details, see [our paper](http://arxiv.org/abs/2507.12927). | |
| ## Limitations | |
| Models are trained for fixed sequence lengths and may perform worse on other lengths or if the test data distribution differs significantly from the training data. |