Skip to content

Pytorch implementation of DeepNovoV2, a state-of-the-art de novo peptide sequencing model.

License

Notifications You must be signed in to change notification settings

rxnataliezhang/DeepNovoV2

 
 

Repository files navigation

DeepNovo-pytorch

The DeepNovoV1 branch contains a pytorch re-implementation of DeepNovo

The DeepNovoV2 branch contains the implementation of our proposed DeepNovoV2 model.

Dependency

python >= 3.6

pytorch >= 1.0

dataclasses

data files

The ABRF DDA spectrums (the data for Table 1 in the original paper) and the default knapsack.npy file could be downloaded here. And the 9 species data could be downloaded here: ftp://massive.ucsd.edu/MSV000081382/peak/DeepNovo/HighResolution/.

It is worth noting that in our implementation we represent training samples in a slightly different format (i.e. peptide stored in a csv file and spectrums stored in mgf files). We also include a script for converting the file format (data_format_converter.py in DeepNovoV2 branch).

usage

first build cython modules

make build

train mode:

make train

denovo mode:

make denovo

evaluate denovo result:

make test

About

Pytorch implementation of DeepNovoV2, a state-of-the-art de novo peptide sequencing model.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 99.7%
  • Makefile 0.3%