Generate chemical structures possibly satisfying desired properties using the inverse QSPR model. It has three reference classes. ENgram is a class for learning the grammar structure of existing chemical strings using an extended N-gram model. QSPRpred contains a simple Bayes regression model to predict properties from structures. SmcChem is a class of the generator of chemical strings from the Inverse-QSPR model. This class has ENgram and QSPRpred class objects inside. The generator is implemented by the Sequential Monte Carlo sampler.
|Date of publication||2016-05-21 12:07:40|
|Maintainer||Hisaki Ikebata <email@example.com>|
engram_5k: ENgram object trained with 5000 SMILES strings
ENgram-class: Extended N-gram model for leraning SMILES strings
Esmi-class: Extend smi class
genENgram: generate SMILES strings from extended N-gram model
get_hiscores: get chemical structures with high QSPR score from SmcChem...
get_score: get QSPR scores for structures in the SmcChem object
get_smiles: get SMILES strings from the SmcChem object
predict: predict properties with forward QSPR model
qspr.data: sample dataset for QSPR modeling
QSPRpred-class: QSPRpredictor class
qsprpred_EG_5k: QSPRpred object trained with 5000 SMILES strings and...
SmcChem-class: SMC chemical generator class
smcexec: modify chemical structures with SMC
trainedSMI: sample dataset for learning with Extended N-gram model
viewstr: view 2D structures from SMILES string vector