Skip to content

Deep Learning-based Gas Chromatographic Retention Index Predictor

Notifications You must be signed in to change notification settings

TomasVrzal/DeepReI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DeepReI

Deep Learning-based Gas Chromatographic Retention Index Predictor

Details are published in: Tomáš Vrzal, Michaela Malečková, Jana Olšovská,DeepReI: Deep learning-based gas chromatographic retention index predictor,Analytica Chimica Acta 1147, 2021, 64 - 71, https://doi.org/10.1016/j.aca.2020.12.043

Development was suported by the Ministry of Agriculture of the Czech Republic within the institutional support MZE-RO1918.

INSTRUCTIONS FOR INSTALLATION:

(If R, keras package and 7-Zip are already installed, step 1 – 4 could be skipped.)

  1. Download and install R (https://www.r-project.org/) and RStudio (https://rstudio.com/products/rstudio/download/).

  2. Download and install 7-Zip (https://www.7-zip.org/).

  3. Run RStudio and install keras package according instructions at https://tensorflow.rstudio.com/guide/keras/.

     install.package(“keras”)
     library(keras)
     install_keras()
    

During the installation, R will ask if miniconda should be installed -> type y to the R console and press Enter button.

  1. Install ChemmineR and ChemmineOB packages by typing the following code in R.

     if (!requireNamespace("BiocManager", quietly = TRUE))
     	install.packages("BiocManager")
    
     BiocManager::install("ChemmineR")
    
     if (!requireNamespace("BiocManager", quietly = TRUE))
     	install.packages("BiocManager")
    
     BiocManager::install("ChemmineOB")
    

  5) Install DeepReI by typing the following code in R.

devtools::install_github("TomasVrzal/DeepReI")

At this moment, the DeepReI uses demo model – prediction will not be performed.

  1. Open 7-Zip File Manager (installed in step #3), find extdata folder in directory where DeepReI package was installed (e.g. Documents\R\win-library\4.0/DeepReI). Click on the model_DeepReI.zip.001 by right mouse button in the 7-Zip File Manager and select „Combine files…“. This procedure creates model_DeepReI.h5 file in zip folder. Unzip the .h5 file and replace the original model_DeepReI.h5 file in extdata folder with unzipped model_DeepReI.h5 file.

After these steps, the DeepReI model is successfully prepared for use. Checking of DeepReI functionality could be carried out by predicting retention indices for 11 compounds in smiles.txt file (saved in DeepReI package directory, see step #6). If all is correct the following table will be obtained after prediction:

SMILES 	RI 
CC 	173.00 
CCC 	316.00 
CCCC 	404.00 
CCCCC 	475.00 
CCCCCC 	601.00 
CCCCCCC 704.00 
CCCCCCCC 807.00 
CCCCCCCCC 894.00 
CCCCCCCCCC 1005.00 
CCCCCCCCCCCCCC 1396.00 
c1ccccc1 661.00 

During this first test of the DeepReI, it is beneficial to see notifications appearing in R console. In some cases, a “Tensorflow not found” error could appear, and, it is possible to fix it by typing the following code to R console (after stopping the DeepReI by clicking on red “stop” button).

tensorflow::install_tensorflow(version = “2.0.0")

INSTRUCTIONS FOR USE:

After successful installation, type following code to the R console.

library(DeepReI)

DeepReI()

DeepReI application will be opened in a web browser. In the SMILES input tab, select file (.csv or .txt) with SMILES in a first column. If the first row represents header of the column, tick the Header box.

At this stage, retention index prediction could be performed by clicking on "Predict Retention Index" button (in Prediction tab).

After few seconds, table with predicted value of retention index together with respective SMILES will appear. This table could be copied and/or exported to Excel format by the appropriate buttons in Export results tab.

About

Deep Learning-based Gas Chromatographic Retention Index Predictor

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages