In jaypat87/HTdescR: Hammett-Taft type sigma descriptor calculator

use htdescbatch to get HT descriptors for the training and test set chemicals

# reading a input csv file with fragments

use clean function to take the output from htdescbatch and convert it into a subset containing only training and test set descriptors

# replacing NA, and subsetting the dataframe to only give what we need

use the qsarml to calculate summary stats for all QSARs

use qsarad for calculating applicability domains for each models

creating a function with these parameters

gethydrolysisrate (inputcsvfile, MCS.cuttoff=0,onlyortho.cutoff=0, regression.method = "SVR", sigma.selection = "A", outputtype = CSV or R dataframe, ...)

get.hydrolysisrate <- function (inputcsvfile = "test.csv", sigma.selection = "A", MCS.cuttoff=0, regression.method = "SVR", outputtype = CSV or R dataframe, ...){ # Calling the helper function to autofill a dataframe which will become test set for our QSARs qsardataframe <- fillqsardataframe (inputcsvfile, sigma.selection = "A", ...)

# Evaluating if the autofilled dataframe does in fact have sigma value higher than the user specified cutoff if (qsardataframe$meta1mcs || meta2mcs || r1taftmcs.... < MCS.cuttoff) { stop ("qsar evaluation terminated since one of the fragments has mcs tanimoto coefficient value below the user specified MCS cuttoff value") }

# Evaluating if the autofilled dataframe does in fact have sigma value higher than the user specified cutoff if (qsardataframe$ortho1mcs || ortho2mcs ||...... < onlyortho.cuttoff) { stop ("qsar evaluation terminated since ortho fragments have mcs tanimoto coefficient value below the user specified MCS cuttoff value") }

# Based on functional group label in qsardataframe, subsetting the master sheet to only keep those variables necassary for a particular qsar if (qsardataframe$funcgroup == "carbamate") { # subset the dataframe to only keep variables used for carbamates and assign it as test set } else if (qsardataframe$funcgroup == "acidester") { # subset the dataframe to only keep variables used for carboxylic acid esters and assign it as test set } else { stop ("given file doesnt contain a functional group with available QSAR model") }

if (regression.method =="SVR") { # support vector regression using e1071 library # call AD function } else if (regression.method == "pls") { # partial least squares using pls library # call AD function } else if (regression.method == "RF") { # random forest regression using randomforest library # call AD function } else if (regression.method == "MLR") { # multiple linear regression using base R linear model functions # call AD function } else { stop ("Specify valid regression.method") }

```

jaypat87/HTdescR documentation built on May 15, 2019, 3:18 p.m.