Copyright 2014 Sebastian Raschka

smilite

smilite is a Python 3 module to download and analyze SMILE strings (Simplified Molecular-Input Line-entry System) of chemical compounds from ZINC (a free database of commercially-available compounds for virtual screening, http://zinc.docking.org)

Sections

Installation

Documentation

Command Line Scripts Examples

Contact

Changelog

Installation

You can use the following command to install PyPrind:
pip install smilite
or
easy_install smilite

Alternatively, you download the package manually from the Python Package Index https://pypi.python.org/pypi/smilite, unzip it, navigate into the package, and use the command:

python3 setup.py install

Documentation

After you installed the smilite module, you can import it in Python via import smilite. The current functions include:

def get_zinc_smile(zinc_id):
    Gets the corresponding SMILE string for a ZINC ID query from
    the ZINC online database. Requires an internet connection.
    Keyword arguments:
        zinc_id (str): A valid ZINC ID, e.g. 'ZINC00029323'
    Returns the SMILE string for the corresponding ZINC ID.
        E.g., 'COc1cccc(c1)NC(=O)c2cccnc2'

def generate_zincid_smile_csv(zincid_list, out_file):
   Generates a CSV file of ZINC_ID,SMILE_string entries by querying the ZINC online
   database.
   Keyword arguments:
        zincid_list (str): Path to a UTF-8 or ASCII formatted file
             that contains 1 ZINC_ID per row. E.g.,
             ZINC0000123456
             ZINC0000234567
             [...]
        out_file (str): Path to a new output CSV file that will be written.

def check_duplicate_smiles(zincid_list, out_file, compare_simplified_smiles=False):
   Scans a ZINC_ID,SMILE_string CSV file for duplicate SMILE strings.
   Keyword arguments:
        zincid_list (str): Path to a UTF-8 or ASCII formatted file that
               contains 1 ZINC_ID per row.
               E.g.,
               ZINC12345678,Cc1ccc(cc1C)OCCOc2c(cc(cc2I)/C=N/n3cnnc3)OC
               ZINC01234567,C[C@H]1CCCC[NH+]1CC#CC(c2ccccc2)(c3ccccc3)O
               [...]
        out_file (str): Path to a new output CSV file that will be written.
        compare_simplified_smiles (bool): If true, SMILE strings will be simplified
               for the comparison.

def simplify_smile(smile_str):
        Simplifies a SMILE string by removing hydrogen atoms (H),
        chiral specifications ('@'), charges (+ / -), '#'-characters,
        and square brackets ('[', ']').
    Keyword Arguments:
        smile_str (str): A smile string, e.g., CC@HNCCS(=O)(=O)[O-])
    Returns a simplified SMILE string, e.g., CC(CCC(=O)NCCS(=O)(=O)O)

Command Line Scripts Examples

If you downloaded the smilite package from https://pypi.python.org/pypi/smilite or https://github.com/rasbt/smilite, you can use the command line scripts I provide in the scripts/ dir.

gen_zincid_smile_csv.py

Generates a ZINC_ID,SMILE_STR csv file from a input file of ZINC IDs. The input file should consist of 1 columns with 1 ZINC ID per row.

Usage:
[shell]>> python3 gen_zincid_smile_csv.py in.csv out.csv

Example:
[shell]>> python3 gen_zincid_smile_csv.py ../examples/zinc_ids.csv ../examples/zid_smiles.csv

Input example file format:

(zinc_ids.csv)

Output example file format:

(zid_smiles.csv)

comp_smile_strings.py

Compares SMILE strings in a 2 column CSV file (ZINC_ID,SMILE_string) to identify duplicates. Generates a new CSV file with ZINC IDs of identified duplicates listed in a 3rd-nth column(s).

Usage:
[shell]>> python3 comp_smile_strings.py in.csv out.csv [simplify]

Example 1:
[shell]>> python3 gen_zincid_smile_csv.py ../examples/zinc_ids.csv ../examples/zid_smiles.csv

Input example file format:

(zid_smiles.csv)

Output example file format 1:

(comp_smiles.csv)

Where
- 1st column: ZINC ID
- 2nd column: SMILE string
- 3rd column: number of duplicates
- 4th-nth column: ZINC IDs of duplicates

Example 2:
[shell]>> python3 comp_smile_strings.py ../examples/zid_smiles.csv ../examples/comp_simple_smiles.csv simplify

Output example file format 2:
(comp_simple_smiles.csv)

Contact

If you have any questions or comments about PyPrind, please feel free to contact me via
eMail: se.raschka@gmail.com
or Twitter: @rasbt

Changelog