Calculate Similarity score
{primary} Since version 21.8.0 there was API upgrade. Use follwing script, if you are using this version or newer, otherwise use the older below.
/*
* Find Similarity Value of multiple selection
*
* Usage:
* 1. Run your Similarity Query - needed, otherwise script will not run
* 2. Run button script
*
* The button script expects that SIMILARITY search
* and Tanimoto metric is used. If you want to use different
* metric, please see all available
* https://docs.chemaxon.com/display/docs/Functions+by+Categories#dissimilarity_functions
*
* Since 20.18.0 API has been changed on the DissimilarityCalculator interface
* https://dl.chemaxon.com/instantjchem/ijc_latest/docs/developer/api/com-chemaxon-ijc-wrappers-dif-api/com/im/df/api/chem/DissimilarityCalculator.html
*
* @author David Pech <dpech@chemaxon.com>
*
*/
import com.im.df.api.*
import com.im.df.api.chem.DissimilarityCalculator
import chemaxon.struc.Molecule
import chemaxon.jep.*
import chemaxon.jep.context.MolContext
import chemaxon.formats.MolImporter;
import com.chemaxon.dif.jchem.calc.*; // since 20.18.0
import chemaxon.sss.search.JChemSearchOptions;
import chemaxon.standardizer.Standardizer;
import com.im.df.api.capabilities.JChemEntityCapability;
init = { widget ->
}
destroy = { widget ->
}
evaluate = { widget ->
def ety = dataTree.rootVertex.entity // assumes you have reference to the data tree
def edp = ety.schema.dataProvider.getEntityDataProvider(ety)
def molFld = ety.fields.items.find { it.name == 'Structure' } // find the structure field
def rs = ety.schema.dataProvider.getDefaultResultSet(dataTree, false, DFEnvironmentRO.DEV_NULL) // find the ResultSet
def rootVS = rs.getVertexState(dataTree.rootVertex) // obtain the VertexState
// define the entity Capability so that we can sniff out the query paramenters
JChemEntityCapability entityCap = DIFUtilities.findCapability(ety, JChemEntityCapability.class);
// // obtain the query paramenters from last used query
boolean isReaction = false
int bitCount = entityCap.getNumberOfOnes();
int bondCount = entityCap.getNumberOfEdges();
int fpLengthInBits = entityCap.getNumberOfInts() * 32;
MarvinStructure queryStructure; // define empty query structure (will be defined in the cycle below)
// obtain the query structure from last used query parameters
List<DFTermExpression> expressions = DIFUtilities.findSimpleFieldUsagesInQuery(rs.getLastExecutedQuery(), molFld);
for (DFTermExpression dFTermExpression : expressions) {
DFOperator operator = dFTermExpression.getOperator();
if (operator instanceof Operators.StructureOperator) {
boolean caseInsensitive = (Boolean) dFTermExpression.getOptions()
// deprecated .get(LegacyConstants.CASE_INSENSITIVE_SEARCH);
.get(Operators.CASE_INSENSITIVE_SEARCH);
List<DFTerm> operands = dFTermExpression.getOperands();
for (DFTerm dFTerm : operands) {
if (dFTerm instanceof DFTermValue && ((DFTermValue) dFTerm).getValue() instanceof MarvinStructure) {
queryStructure = ((DFTermValue) dFTerm).getValue(); // assign the query structure value to a variable
}
}
}
}
List ids = rootVS.getSelectedRowsIds() // get the IDs of the selection
// obtain the structure field from Entity
// expects Structure Entity type
DFField structureField = ety.getStructureField()
Set<DFField> set = new HashSet<DFField>();
set.add(structureField);
// obtain the data
// method expects list of IDs and a HashSet of the fields to obtain
// as we defined above, the HashSet contains only Structure field
Map rows = rootVS.getData(ids, set, DFEnvironmentRO.DEV_NULL)
// Create the dissimilarity calculator and
// set other needed parameters for the calculation
JChemEntityCapability jchem = DIFUtilities.findCapability(structureField.getEntity(), JChemEntityCapability.class); //Since 20.18.0 with new implementation
JChemSearchOptions jcso = new JChemSearchOptions(JChemSearchOptions.SIMILARITY); // set SIMILARITY search type
jcso.setDissimilarityMetric("TANIMOTO"); // use TANIMOTO metric
Molecule query = queryStructure.getNative() // obtain the chemaxon.struct.Molecule instance and set it as query
Standardizer noSt = new Standardizer("<StandardizerConfiguration><Actions></Actions></StandardizerConfiguration>")
// create Dissimilarity calculator with all needed paramenters
//Replaced by JChemDissimilarityCalculator since 20.18.0 *DissimilarityCalculator dissimilarity = new DissimilarityCalculator(isReaction, query, jcso, bitCount, bondCount, fpLengthInBits, noSt);
JChemDissimilarityCalculator dissimilarity = new JChemDissimilarityCalculator(query, jcso, noSt, jchem);
// cycle through the data, for each row obtain the structure
// convert it to chemaxon.struct.Molecule object
// and parse it to dissimilarity value calculator
for (row in rows.values()) {
MarvinStructure mol = row[molFld.id]
Molecule cxnMol = mol.getNative()
//Replaced by JChemDissimilarityCalculator since 20.18.0 *def dissimilarityValue = dissimilarity.computeDissimilarity(cxnMol)
def dissimilarityValue = dissimilarity.calculateDissimilarity(cxnMol) // caluclate Disimilarity score for selected molecule
def similarityValue = 1 - dissimilarityValue // calculate the Similarity value
println "Similarity Value is ${similarityValue}"
}
}
on_change = { widget, button ->
}
Until version 20.18.0 use this version of script below
/*
* Find Similarity Value of multiple selection
*
*
* Usage:
* 1. Run your Similarity Query - needed, otherwise script will not run
* 2. Run button script
*
* The button script expects that SIMILARITY search
* and Tanimoto metric is used. If you want to use different
* metric, please see all available
* https://docs.chemaxon.com/display/docs/Functions+by+Categories#dissimilarity_functions
*
* @author David Pech <dpech@chemaxon.com>
*/
import com.im.df.api.*
import com.im.df.api.chem.DissimilarityCalculator
import chemaxon.struc.Molecule
import chemaxon.jep.*
import chemaxon.jep.context.MolContext
import chemaxon.formats.MolImporter;
import chemaxon.sss.search.JChemSearchOptions;
import chemaxon.standardizer.Standardizer;
import com.im.df.api.capabilities.JChemEntityCapability;
init = { widget ->
}
destroy = { widget ->
}
evaluate = { widget ->
def ety = dataTree.rootVertex.entity // assumes you have reference to the data tree
def edp = ety.schema.dataProvider.getEntityDataProvider(ety)
def molFld = ety.fields.items.find { it.name == 'Structure' } // find the structure field
def rs = ety.schema.dataProvider.getDefaultResultSet(dataTree, false, DFEnvironmentRO.DEV_NULL) // find the ResultSet
def rootVS = rs.getVertexState(dataTree.rootVertex) // obtain the VertexState
// define the entity Capability so that we can sniff out the query paramenters
JChemEntityCapability entityCap = DIFUtilities.findCapability(ety, JChemEntityCapability.class);
// // obtain the query paramenters from last used query
boolean isReaction = false
int bitCount = entityCap.getNumberOfOnes();
int bondCount = entityCap.getNumberOfEdges();
int fpLengthInBits = entityCap.getNumberOfInts() * 32;
MarvinStructure queryStructure; // define empty query structure (will be defined in the cycle below)
// obtain the query structure from last used query parameters
List<DFTermExpression> expressions = DIFUtilities.findSimpleFieldUsagesInQuery(rs.getLastExecutedQuery(), molFld);
for (DFTermExpression dFTermExpression : expressions) {
DFOperator operator = dFTermExpression.getOperator();
if (operator instanceof Operators.StructureOperator) {
boolean caseInsensitive = (Boolean) dFTermExpression.getOptions()
.get(LegacyConstants.CASE_INSENSITIVE_SEARCH);
List<DFTerm> operands = dFTermExpression.getOperands();
for (DFTerm dFTerm : operands) {
if (dFTerm instanceof DFTermValue && ((DFTermValue) dFTerm).getValue() instanceof MarvinStructure) {
queryStructure = ((DFTermValue) dFTerm).getValue(); // assign the query structure value to a variable
}
}
}
}
List ids = rootVS.getSelectedRowsIds() // get the IDs of the selection
// obtain the structure field from Entity
// expects Structure Entity type
DFField structureField = ety.getStructureField()
Set<DFField> set = new HashSet<DFField>();
set.add(structureField);
// obtain the data
// method expects list of IDs and a HashSet of the fields to obtain
// as we defined above, the HashSet contains only Structure field
Map rows = rootVS.getData(ids, set, DFEnvironmentRO.DEV_NULL)
// Create the dissimilarity calculator and
// set other needed parameters for the calculation
JChemSearchOptions jcso = new JChemSearchOptions(JChemSearchOptions.SIMILARITY); // set SIMILARITY search type
jcso.setDissimilarityMetric("TANIMOTO"); // use TANIMOTO metric
Molecule query = queryStructure.getNative() // obtain the chemaxon.struct.Molecule instance and set it as query
Standardizer noSt = new Standardizer("<StandardizerConfiguration><Actions></Actions></StandardizerConfiguration>")
// create Dissimilarity calculator with all needed paramenters
DissimilarityCalculator dissimilarity = new DissimilarityCalculator(isReaction, query, jcso, bitCount, bondCount, fpLengthInBits, noSt);
// cycle through the data, for each row obtain the structure
// convert it to chemaxon.struct.Molecule object
// and parse it to dissimilarity value calculator
for (row in rows.values()) {
MarvinStructure mol = row[molFld.id]
Molecule cxnMol = mol.getNative()
def dissimilarityValue = dissimilarity.computeDissimilarity(cxnMol) // caluclate Disimilarity score for selected molecule
def similarityValue = 1 - dissimilarityValue // calculate the Similarity value
println "Similarity Value is ${similarityValue}"
}
}
on_change = { widget, button ->
}