bpnext 0.0.1

yet another BPNet implementation in PyTorch

Next-generation BPNet models

[!WARNING] This is an experimental playground for BPNet-like models.

BPNet is a family of base-resolution DNA sequence models.

This repository is an experimental playground for different models stemming from the original BPNet architecture.

Model implementations in this repository are heavily based on, and would not be possible without, their original implementations, which are linked below.

Models

BPNet

from bpnext.models import BPNet

BPNet is a model for predicting base-resolution transcription factor binding profiles from DNA sequences. BPNet models are trained using ChIP-nexus or ChIP-seq data.

• The model is described in detail in the publication Avsec et al., 2021.

• The model was originally implemented in TensorFlow by zigaavsec: bpnet.

ChromBPNet

from bpnext.models import ChromBPNet

ChromBPNet is a model for predicting base-resolution chromatin accessibility profiles from DNA sequences. ChromBPNet models are trained using ATAC-seq or DNase-seq data.

• The model and its applications are described in detail in the preprint Pampari et al., 2024.

• The model was originally implemented in TensorFlow by panushri25: chrombpnet.

• Object-oriented implementation of ChromBPNet has been later written by jmschrei in PyTorch: dragonnfruit.

DragoNNFruit

from bpnext.models import DragoNNFruit

DragoNNFruit is a model for predicting base-resolution chromatin accessibility profiles from DNA sequences and cell states. The model has been designed to be trained on single-cell ATAC-seq data and multiome data.

This model is currently under development.

• The model has been originally implemented in PyTorch by jmschrei: dragonnfruit.

Citations

If you use this code and these models in your research, please consider citing this repository as well as the following publications:

BPNet:

@article{avsec2021base,
  title={Base-resolution models of transcription-factor binding reveal soft motif syntax},
  author={Avsec, {\v{Z}}iga and Weilert, Melanie and Shrikumar, Avanti and Krueger, Sabrina and Alexandari, Amr and Dalal, Khyati and Fropf, Robin and others},
  journal={Nature genetics},
  volume={53},
  number={3},
  pages={354--366},
  year={2021},
  publisher={Nature Publishing Group}
}

ChromBPNet:

@article{pampari2024chrombpnet,
  title={ChromBPNet: bias factorized, base-resolution deep learning models of chromatin accessibility reveal cis-regulatory sequence syntax, transcription factor footprints and regulatory variants},
  author={Pampari, Anusri and Shcherbina, Anna and Kvon, Evgeny Z and Kosicki, Michael and Nair, Surag and Kundu, Soumya and Kathiria, Arwa S and others},
  journal={bioRxiv},
  year={2024},
  publisher={Cold Spring Harbor Laboratory}
}

License

This project is licensed under the MIT License. See the LICENSE file for details.