README.md

In jack-davison/chemdesr: Obtain Data from the ChemDes Platform

{chemdesr} 📦

Functions used to scrape chemical information from the ChemDes platform, more specifically to assist with running the model found at https://dan-wacl.shinyapps.io/rie_shiny_app/.

To install, type the below code into your R console:

devtools::install_github('jack-davison/chemdesr')

Features

There are two key functions in chemdesr: read_chemdes: Provided a simplified molecular-input line-entry system ("SMILES") string and a specific descriptor (e.g. Chemopy, CDK, etc.), scrapes the ChemDes platform and returns a wide data frame of the relevant chemical information. read_full_chemdes: A wrapper around read_chemdes which automatically reads in Chemopy, CDK, RDKit, Pybel, and PaDEL data for a given SMILES string. There is an option to make this function "chatty" if you would like it to update you on its progress. * map_chemdes: A wrapper around read_full_chemdes to make it easier to iterate over multiple compounds.

There are also example data: * chem_smiles: A 42x2 dataframe containing 42 example compounds and their SMILES strings. Use this to practice with the chemdesr functions.

Example Workflows

It is very simple to scrape a single compound using chemdesr. Your workflow may look something like this:

single_smiles = "OC(C)(C(CC(O)=O)CC(O)=O)C"

# Read a single descriptor (in this case Chemopy)
chemdesr::read_chemdes(single_smiles, desc = "chemopy")

# Read multiple descriptors
chemdesr::read_full_chemdes(single_smiles, chatty = T)

For multiple compounds, I'd suggest using a map function from purrr. For many compounds, I would recommend enabling the "chatty" option to be kept abreast of the function's progress. If you aren't confident with purrr, there is an in-built map_chemdes function to do this for you.

my_smiles = c("OC(C)(C(CC(O)=O)CC(O)=O)C", 
              "C(C=CC1)=CC=1C(=O)O", 
              "CC1(C)OC(=O)CC1CC(O)=O")

# Do it yourself
purrr::map_dfr(.x = my_smiles, .f = chemdesr::read_full_chemdes, chatty = T)

# Built-in mapping (equivalent to the above)
chemdesr::map_chemdes(my_smiles, chatty = T)

With a data frame of SMILES strings, one could reobtain the non-SMILE information through joining functions.

compounds = chemdesr::chem_smiles %>% head(3)

comp_desc = map_dfr(.x = pull(compounds, smile), .f = chemdesr::read_full_chemdes, chatty = T)

left_join(compounds, comp_desc) 

jack-davison/chemdesr documentation built on Sept. 24, 2022, 10:22 a.m.