scripts for structural interpretation of QSAR models
Python
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore
README.md
calc_contributions.py
fragmentation.py

README.md

spci-ext

Scripts of generation of counter-fragments and calculation of fragments contributions in order to facilitate structural interpretation of QSAR models.

fragmentation.py

Generates counter-fragments for calculation of descriptors and QSAR prediction.

There are three required options:

  • input sdf or smiles files (-i option)
  • output sdf or smiles file (-o)
  • fragmentation scheme chosen across available ones (-f)

If compounds have explicit hydrogens and output is an sdf file:

  1. it is recommended to remove detached single hydrogens remaining after removal of a counter-fragment (-x). In such a way you estimate contribution of the fragment with all attached hydrogens.
  2. if software cannot calculate descriptors for compounds with free valence use option -a to automatically add hydrogens to cap free valence

If used descriptor software cannot calculate descriptors for multi-component counter-fragments just remove them from the output (-z). In such a case you will not be able to estimate contributions of scaffolds/linkers, only terminal fragments.

calc_contributions.py

Calculates contributions from QSAR predictions for fragments generated by the previous script. Input file should contain two columns with name and predicted value separated by space or tab.

Help

Call scripts with -h key.

Visualization of results

I implemented visualisation with R package rspci: https://github.com/DrrDom/rspci To install it run in R console: ''' devtools::install_github("DrrDom\rspci") ''' To load data: ''' df <- rspci::load_data_ext(file_name_with_contributions.txt) ''' and then as described in rspci package.

License

GPLv3