TwinC is a Python package for training, inference, and interpretation of trans-3D genome folding in humans. TwinC uses a Convolutional Neural Network model that predicts contact between two trans genomic loci from nucleotide sequences. The model takes two 100 kbp nucleotide sequences as input and treats the task of predicting trans contacts as a classification task. To reproduce analyses performed in the Jha et al., 2024, please check out TwinC 2024 manuscript repository.
python -m pip install git+https://github.com/Noble-Lab/twinc
Training a new TwinC model
twinc_train --config_file configs/heart_left_ventricle.yml
Inference on TwinC model
twinc_test --config_file configs/heart_left_ventricle.yml
Jha A, Hristov BH, Wang X, Wang S, Greenleaf WJ, Kundaje A, Lieberman Aiden E, Bertero A, Noble WS. Prediction and functional interpretation of inter-chromosomal genome architecture from DNA sequence with TwinC. bioRxiv (2024): 2024-09. doi: https://doi.org/10.1101/2024.09.16.613355
We welcome any bug reports, feature requests or other contributions. Please submit a well-documented report on our issue tracker. For substantial changes, please fork this repo and submit a pull request for review.
See CONTRIBUTING.md for additional details.
You can find official releases here.