true
The goal of fcmsafety is to assign toxicity levels for compounds from food contact materials. It takes into account the toxicity data from: 1. Substances of Very High Concern (SVHC) from ECHA (https://echa.europa.eu/candidate-list-table); 2. Carcinogenic, Mutagenic, and Reprotoxic (CMR) from the Classification, Labelling, and Packaging (CLP) regulation (https://echa.europa.eu/information-on-chemicals/annex-vi-to-clp) ; 3. Carcinogenic substances from IARC (https://monographs.iarc.who.int/list-of-classifications); 4. Endocrine Disrupting Chemicals (EDC) from The International Panel on Chemical Pollution (IPCP) commissioned by UN Environment (https://www.unep.org/explore-topics/chemicals-waste/what-we-do/emerging-issues/scientific-knowledge-endocrine-disrupting); 5. Specific Migration Limit (SML) from EU 10/2011 regulation (https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:02011R0010-20200923&qid=1636402301680&from=en) 6. Specific Migration Limit (SML) from China GB 9685 regulation (Provided by IQTC).
The toxicity level assignment follows the rules listed in the inst/toxicity_levels.png (don’t know why it is not rendered in this readme file)
Please bear in mind that the toxicity level assignment here is on the basis of individual compounds, whilst substances listed in the aforementioned databases could be groups of compounds. These groups of substances will not be considered as they not are specific in terms of chemical structures and no “InChIKey” can be retrieved from Pubchem, which is used for matching. For example, nonylphenols are a family of closely related organic compounds composed of phenol bearing a 9 carbon-tail, and they could be overlooked unless any database has numerated some specific structures in their lists.
You can install the development version of fcmsafety from GitHub with the code listed below.
install.packages("remotes")
remotes::install_github("QizhiSu/fcmsafety")library(fcmsafety)
library(dplyr)You have to first update all the databases and load them into the global environment.
# You don't have to update all the databases as we have prepared them (updated
# on 2021/11/11) and can be used directly. In case you want to have the latest
# databases (if you find any update of these databases after 2021/11/11), please
# run update_databases(), it might take long time as retrieving meta data for
# all these databases is time-consuming.
update_databases()
# If you have updated the databases yourself, please run:
load_databases(use_default = FALSE)
# Otherwise
load_databases()# Please enter the path of your data, e.g., "D:/my data/mydata.xlsx".
data <- rio::import("D:/my data/mydata.xlsx")Given that your data have at least SMILES and CAS numbers, you can use the following code to process your data. Otherwise, you need to firstly extract metadata from Pubchem using the labtools.
For compounds that do not present in any of the databases we consider here, we can use Toxtree (http://toxtree.sourceforge.net/) to predict its Cramer rules level which will be used for assigning toxicity level in the following step. We now can export a file that can be used by Toxtree for batch processing.
# This code will generate a file name "for_toxtree.csv"
export4toxtree(data, cas_col = 1, name_col = 2)
# Of course, you can also specify the path and name of the output file
export4toxtree(data, cas_col = 1, name_col = 2, output = "D:/data/mydata_for_toxtree.csv"){width=500px}
{width=500px}
{width=500px}
Normally, this step is quite fast, but we noticed that polyethylene glycols might take quite long time.
This step requires the results from Toxtree (assuming you have the result in “D:/data/mydata_toxtree_results.csv”).
data_complete <- data %>%
assign_toxicity(toxtree_result = "D:/data/mydata_toxtree_results.csv")rio::export(data_complete, "D:/data/mydata_result.xlsx")