jinysun/fastei

Ultra-fast and Accurate Spectrum Matching

FastEI is an ultra-fast and accurate spectrum matching method, proposed to improve accuracy by Word2vec-based spectrum embedding and boost the speed using hierarchical navigable small world graph (HNSW)

Motivation

The in-silico library with predicted spectra of large-scale molecules can extend the chemical space and increase the coverage immensely when compared with experimental libraries (e.g., NIST 2017 and MassBank libraries). How to rapidly search an in-silico library of millions or even tens of millions of spectra while ensuring the accuracy of molecular identification is a new challenge. In this work, a million-molecule scale in-silico library has been builded and an ultra-fast and accurate search method has been developed (FastEI).

Depends

Anaconda for Python 3.7

conda install -c conda-forge gensim

conda install -c rdkit rdkit

conda install -c conda-forge hnswlib

Installation

The current install version of FastEI only supports Windows 64-bit version. It has been test on windows 7, windows 10.

Install software: FastEI-GUI-1.0.0-Windows.exe

Note: All the database, HNSW index, Word2vec model and query spectra examples are included in the installation package. So there are four files needed to be downloaded, including FastEI-GUI.exe, FastEI-GUI-1.bin, FastEI-GUI-2.bin, and FastEI-GUI-3.bin. It will take some time to install. Please be patient.

Development version

1. Install Anaconda or Miniconda

2. Install Git

3. Open commond line, clone the repository and enter:

   git clone https://github.com/Qiong-Yang/FastEI.git
   cd FastEI
   

4. Create environment and install dependency with the following commands :

   conda env create -f FastEI.yml conda activate FastEI

5. Run FastEI.py:

   cd GUI/ui
   python FastEI.py
   

Usage

Software

The video for using the FastEI is available at the video folder.

FastEI-GUI.mov

For the details on how to use FastEI, please check Ducomentation.

Code

Database, Word2vec model and HNSW index download:

Please put IN_SILICO_LIBRARY.db , references_index.bin and references_word2vec.model into data directory.

Take example.py （example.ipynb）as an example for molecular identification. If you want to identify molecules based on your spectra, please put your spectra files in to spectra directory and run test.py.

Contact

Yang qiong

E-mail: [email protected]