An open API service indexing awesome lists of open source software.

https://github.com/aspirincode/deepfrag

A pre-train and fine-tune method for predicting MS/MS spectra for a specific type of MS instrument
https://github.com/aspirincode/deepfrag

Last synced: 12 months ago
JSON representation

A pre-train and fine-tune method for predicting MS/MS spectra for a specific type of MS instrument

Awesome Lists containing this project

README

          

## DeepFrag
***
A pre-train and fine-tune method for predicting MS/MS spectra for a specific type of MS instrument.



### Depends:
Anaconda for python >= 3.6
# RDKit
conda install -c rdkit rdkit
# Keras
conda install keras
# pycdk
pip install git+git://github.com/hcji/pycdk@master
# pyCFMID (optional, just used for comparing with CFM-ID)
pip install git+git://github.com/hcji/PyCFMID@master

### Scripts:
All scripts used for this work are included in the **Scripts** directory.
**Scripts/train_model.py** is the scrpit which is used for pre-train the model with CFM predicted spectra. The spectra are available at [MINE](http://minedatabase.mcs.anl.gov).
**Scripts/transfer_model_RIKEN_PlaSMA.py** is the script which is used for fine-tune the model with RIKEN PlaSMA database.
**Scripts/example1.py** and **Scripts/example2.py** are two examples from the test dataset.
**Scripts/annotation_example.py** and **Scripts/annotation_example2.py** are two examples explaining how to annotate the predicted spectra.

### Graphical Interfaces
This repo doesn't contain the model. One can download the whole DeepFrag at [url](https://figshare.com/articles/DeepFrag_zip/8323568). Then use the graphical interfaces with

cd ../DeepFrag
python DeepFragApp.py