Skip to content

Latest commit

 

History

History
44 lines (36 loc) · 2.79 KB

README.md

File metadata and controls

44 lines (36 loc) · 2.79 KB

VIDEAN (Visual and Interactive Descriptor Analysis)

Available on: http://lidecc.cs.uns.edu.ar/VIDEAN/

Main directories and files of the tool

  • cgi-bin: contains a python script to run external functions of Weka.
  • cgi-data: contains external data files with results of run the python script.
  • css : contains the cascading style sheets.
  • d3 : library d3.js (http://d3js.org/).
  • data : contains the data files used in the examples.
  • javascript : javascript files
    • VIDEAN_View1: load the data for the "Molecular descriptors and their correlations" section.
    • VIDEAN_View2: load the data for the "Co-occurrence of molecular descriptors" section.
    • VIDEAN_View3: load the data for the "Models and descriptors" section.
    • functions_view1: implementation of the functions used to visualize and interact with data loaded by VIDEAN_View1.
    • functions_view2: implementation of the functions used to visualize and interact with data loaded by VIDEAN_View2.
    • functions_histogram_scatterplot: implementation of the functions used to visualize the charts of section "Histograms / Scatterplot".
    • load_example: implements the data loading of the tool.

Preprocessing for VIDEAN

dataLoadJSON(file_desc_value,file_desc_names,file_desc_names_complete,file_property_values,model_file,property_name)

  • file_desc_value: csv file that contains the selected descriptors values. Each column contains the values for a descriptor.
  • file_desc_names: excel file containing the selected descriptors names.
  • file_desc_names_complete: excel file containing the names of all the descriptors were admitted to descriptor selection software.
  • file_property_values: csv file that contains the property values.
  • model_file: text file containing the models with descriptors selected in each of them.
  • property_name: the property name

All files generated by the script 'dataLoadJson', which will be used by the visualization tool VIDEAN, are saved in "./output data/". You have to provide the url of this folder to use VIDEAN with your own data.

Important:

  1. you need the jsonlab tool provided in the './jsonlab/' directory. See README.txt (II. Installation) therein.
  2. you need the file server.php in the ./output_data/ directory, which acts as a jsonp server for cross-domain requests.

Example: dataLoadJSON('./input data/selected_descriptors_values.csv','./input data/selected_descriptors_names','./input data/descriptors_names','./input data/property_values.csv','./input data/models.txt','logPliver')

María Jimena Martínez, Ignacio Ponzoni, Mónica F. Díaz, Gustavo E. Vazquez, Axel J. Soto: "Visual Analytics in Cheminformatics: User-Supervised Descriptor Selection for QSAR Methods". Journal of Cheminformatics, 7:39, 2015.