/DeepSequence-torch

A pytorch version of DeepSequence

Primary LanguagePythonMIT LicenseMIT

PyTorch version of DeepSequence

Installation

Create environment

conda env create -f environment.yml

Usage

Activate environment

conda activate deepseq-env

Train DeepSequence VAE model from alignment

python train.py. -a "./datasets/BLAT_ECOLX_hmmerbit_plmc_n5_m30_f50_t0.2_r24-286_id100_b105.a2m" \
  -o ./datasets

Predict with a DeepSequence VAE model

python predict.py -a "datasets/BLAT_ECOLX_hmmerbit_plmc_n5_m30_f50_t0.2_r24-286_id100_b105.a2m" \
  -m "datasets/model_149000_391.404.pt" --n_iter 500

Any contributions are welcome!

Note: DeepSequence is not suitable for online use, as it requires n_iter (usually 500) of predictions to get an accurate delta_elbo.

Special thanks

  • Original repository: DeepSequence implemented with Theano.

License: MIT License