migrated to new SiNGA version

fkaiserbio · Jul 20, 2018 · a4e52ce · a4e52ce
1 parent 5de48d2
commit a4e52ce
Show file tree

Hide file tree

Showing 31 changed files with 170 additions and 116 deletions.
diff --git a/.idea/artifacts/fit3d_web_war.xml b/.idea/artifacts/fit3d_web_war.xml
diff --git a/.idea/artifacts/fit3d_web_war_exploded.xml b/.idea/artifacts/fit3d_web_war_exploded.xml
diff --git a/README.md b/README.md
@@ -9,7 +9,7 @@ Fit3D is a software for template-based and template-free detection of substructu
 ## Background
 The biological function of proteins and nucleic acids, such as riboswitches or ribozymes, relies on the correct arrangement of small substructural units to catalyze substrates, bind ligands, or to 
 preserve an ordered state. These molecular building blocks have evolved to retain similar interaction patterns and geometrical features to ensure functionality. While some of these patterns are 
-reflected in recurring sequence motifs, evolutionarily remote proteins may only share a small set of similar structural motifs, which are not obvious on sequence level. The Fit3D algorithm is a 
+reflected in recurring sequence motifs, evolutionarily remote proteins may only share a small set of similar structural motifs, which are not obvious at sequence level. The Fit3D algorithm is a 
 highly accurate search engine to detect substructure similarities based on a given template. Additionally, it uses the [mmm](https://github.com/fkaiserbio/mmm) framework
 to discover structural motifs without the need for a template.
 

diff --git a/fit3d-cli/fit3d-cli.iml b/fit3d-cli/fit3d-cli.iml
@@ -15,9 +15,9 @@
     <orderEntry type="library" name="R User Library" level="project" />
     <orderEntry type="library" name="R Skeletons" level="application" />
     <orderEntry type="library" name="Maven: commons-cli:commons-cli:1.4" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-structure:0.3.3" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-mathematics:0.3.3" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-features:0.3.3" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-structure:0.4.0" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-mathematics:0.4.0" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-features:0.4.0" level="project" />
     <orderEntry type="library" name="Maven: tec.uom:uom-se:1.0.8" level="project" />
     <orderEntry type="library" name="Maven: javax.measure:unit-api:1.0" level="project" />
     <orderEntry type="library" name="Maven: tec.uom.lib:uom-lib-common:1.0.2" level="project" />
@@ -29,8 +29,8 @@
     <orderEntry type="library" scope="RUNTIME" name="Maven: org.apache.logging.log4j:log4j-api:2.6.2" level="project" />
     <orderEntry type="library" name="Maven: org.rcsb:mmtf-api:1.0.7" level="project" />
     <orderEntry type="library" name="Maven: org.slf4j:slf4j-api:1.7.21" level="project" />
-    <orderEntry type="library" name="Maven: bio.fkaiser.mmm:mmm-core:0.1.2" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-core:0.3.3" level="project" />
+    <orderEntry type="library" name="Maven: bio.fkaiser.mmm:mmm-core:0.2.0" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-core:0.4.0" level="project" />
     <orderEntry type="library" name="Maven: com.fasterxml.jackson.core:jackson-core:2.8.7" level="project" />
     <orderEntry type="library" name="Maven: com.fasterxml.jackson.core:jackson-annotations:2.8.7" level="project" />
     <orderEntry type="library" name="Maven: com.fasterxml.jackson.core:jackson-databind:2.8.7" level="project" />

diff --git a/fit3d-cli/pom.xml b/fit3d-cli/pom.xml
@@ -5,7 +5,7 @@
     <modelVersion>4.0.0</modelVersion>
 
     <properties>
-        <fit3d-cli.version>0.2.0</fit3d-cli.version>
+        <fit3d-cli.version>0.3.0</fit3d-cli.version>
     </properties>
 
     <groupId>bio.fkaiser.fit3d</groupId>
@@ -108,14 +108,14 @@
             <version>1.4</version>
         </dependency>
         <dependency>
-            <groupId>de.bioforscher.singa</groupId>
+            <groupId>bio.singa</groupId>
             <artifactId>singa-structure</artifactId>
-            <version>0.3.3</version>
+            <version>0.4.0</version>
         </dependency>
         <dependency>
             <groupId>bio.fkaiser.mmm</groupId>
             <artifactId>mmm-core</artifactId>
-            <version>0.1.2</version>
+            <version>0.2.0</version>
         </dependency>
         <dependency>
             <groupId>me.tongfei</groupId>

diff --git a/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/Fit3DCommandLine.java b/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/Fit3DCommandLine.java
@@ -7,24 +7,24 @@
 import bio.fkaiser.mmm.model.mapping.rules.ChemicalGroupsMappingRule;
 import bio.fkaiser.mmm.model.mapping.rules.FunctionalGroupsMappingRule;
 import bio.fkaiser.mmm.model.plip.PlipPostRequest;
-import de.bioforscher.singa.core.utility.Resources;
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.representations.RepresentationSchemeType;
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.statistics.FofanovEstimation;
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.statistics.StarkEstimation;
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.statistics.StatisticalModel;
-import de.bioforscher.singa.structure.model.families.AminoAcidFamily;
-import de.bioforscher.singa.structure.model.identifiers.LeafIdentifier;
-import de.bioforscher.singa.structure.model.interfaces.Atom;
-import de.bioforscher.singa.structure.model.interfaces.LeafSubstructure;
-import de.bioforscher.singa.structure.model.interfaces.Structure;
-import de.bioforscher.singa.structure.model.oak.StructuralEntityFilter;
-import de.bioforscher.singa.structure.model.oak.StructuralEntityFilter.AtomFilter;
-import de.bioforscher.singa.structure.model.oak.StructuralMotif;
-import de.bioforscher.singa.structure.parser.pdb.rest.cluster.PDBSequenceCluster;
-import de.bioforscher.singa.structure.parser.pdb.rest.cluster.PDBSequenceCluster.PDBSequenceClusterIdentity;
-import de.bioforscher.singa.structure.parser.pdb.structures.SourceLocation;
-import de.bioforscher.singa.structure.parser.pdb.structures.StructureParser;
-import de.bioforscher.singa.structure.parser.pdb.structures.StructureParserException;
+import bio.singa.core.utility.Resources;
+import bio.singa.structure.algorithms.superimposition.fit3d.representations.RepresentationSchemeType;
+import bio.singa.structure.algorithms.superimposition.fit3d.statistics.FofanovEstimation;
+import bio.singa.structure.algorithms.superimposition.fit3d.statistics.StarkEstimation;
+import bio.singa.structure.algorithms.superimposition.fit3d.statistics.StatisticalModel;
+import bio.singa.structure.model.families.AminoAcidFamily;
+import bio.singa.structure.model.identifiers.LeafIdentifier;
+import bio.singa.structure.model.interfaces.Atom;
+import bio.singa.structure.model.interfaces.LeafSubstructure;
+import bio.singa.structure.model.interfaces.Structure;
+import bio.singa.structure.model.oak.StructuralEntityFilter;
+import bio.singa.structure.model.oak.StructuralEntityFilter.AtomFilter;
+import bio.singa.structure.model.oak.StructuralMotif;
+import bio.singa.structure.parser.pdb.rest.cluster.PDBSequenceCluster;
+import bio.singa.structure.parser.pdb.rest.cluster.PDBSequenceCluster.PDBSequenceClusterIdentity;
+import bio.singa.structure.parser.pdb.structures.SourceLocation;
+import bio.singa.structure.parser.pdb.structures.StructureParser;
+import bio.singa.structure.parser.pdb.structures.StructureParserException;
 import org.apache.commons.cli.*;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
@@ -78,6 +78,8 @@ public class Fit3DCommandLine {
     private int numberOfThreads = DEFAULT_NUMBER_OF_THREADS;
     private Path outputDirectoryPath;
     private ItemsetMinerConfiguration<String> itemsetMinerConfiguration;
+    private boolean distanceFiltering;
+    private Double filterThreshold;
 
     public Fit3DCommandLine(Fit3DMode mode, CommandLine commandLine) throws ParseException {
         this.mode = mode;
@@ -471,6 +473,17 @@ private void initializeTemplateBasedParameters() throws Fit3DCommandLineExceptio
         if (commandLine.hasOption('P')) {
             createStatisticalModel();
         }
+
+        // check if distance filtering is enabled
+        if (commandLine.hasOption('y')) {
+            try {
+                distanceFiltering = true;
+                filterThreshold = Double.valueOf(commandLine.getOptionValue('y'));
+            } catch (NumberFormatException e) {
+                logger.error("failed to parse distance filtering threshold", e);
+                throw new Fit3DCommandLineException("Initialization of parameters failed.");
+            }
+        }
     }
 
     private void createStatisticalModel() throws ParseException {
@@ -607,6 +620,10 @@ public Double getDistanceTolerance() {
         return distanceTolerance;
     }
 
+    public Double getFilterThreshold() {
+        return filterThreshold;
+    }
+
     public ItemsetMinerConfiguration<String> getItemsetMinerConfiguration() {
         return itemsetMinerConfiguration;
     }
@@ -651,6 +668,10 @@ public Path getTargetListPath() {
         return targetListPath;
     }
 
+    public boolean isDistanceFiltering() {
+        return distanceFiltering;
+    }
+
     public boolean isEcMapping() {
         return ecMapping;
     }

diff --git a/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/Fit3DTemplateBasedCommandLineRunner.java b/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/Fit3DTemplateBasedCommandLineRunner.java
@@ -1,11 +1,11 @@
 package bio.fkaiser.fit3d.cli;
 
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.Fit3D;
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.Fit3DBuilder;
-import de.bioforscher.singa.structure.algorithms.superimposition.fit3d.Fit3DMatch;
-import de.bioforscher.singa.structure.model.interfaces.LeafSubstructureContainer;
-import de.bioforscher.singa.structure.parser.pdb.structures.StructureParser;
-import de.bioforscher.singa.structure.parser.pdb.structures.StructureParserOptions;
+import bio.singa.structure.algorithms.superimposition.fit3d.Fit3D;
+import bio.singa.structure.algorithms.superimposition.fit3d.Fit3DBuilder;
+import bio.singa.structure.algorithms.superimposition.fit3d.Fit3DMatch;
+import bio.singa.structure.model.interfaces.LeafSubstructureContainer;
+import bio.singa.structure.parser.pdb.structures.StructureParser;
+import bio.singa.structure.parser.pdb.structures.StructureParserOptions;
 import me.tongfei.progressbar.ProgressBar;
 import me.tongfei.progressbar.ProgressBarStyle;
 import org.slf4j.Logger;
@@ -151,6 +151,11 @@ private void run() throws Fit3DCommandLineException {
             parameterStep.mapUniProtIdentifiers();
         }
 
+        // set environment filtering
+        if(commandLine.isDistanceFiltering()){
+            parameterStep.filterEnvironments(commandLine.getFilterThreshold());
+        }
+
         Timer timer = null;
         ProgressTask progressTask = null;
         if (multiParser != null) {

diff --git a/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/TemplateBasedCommandLineOptions.java b/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/TemplateBasedCommandLineOptions.java
@@ -93,6 +93,11 @@ private void generateOptions() {
         rmsd.setArgs(1);
         options.addOption(rmsd);
 
+        Option distanceFiltering = new Option("y","filter local environments based on distance thresholds (default: disabled)");
+        distanceFiltering.setLongOpt("distance-filtering");
+        distanceFiltering.setArgs(1);
+        options.addOption(distanceFiltering);
+
         // motif extraction
         Option motifExtraction = new Option("X",
                                             "extract motif from structure input (-m) following the syntax [chain]-[residue type][residue number]_... (e.g. A-E651_A-D649_A-T177)\n" +

diff --git a/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/TemplateFreeCommandLineOptions.java b/fit3d-cli/src/main/java/bio/fkaiser/fit3d/cli/TemplateFreeCommandLineOptions.java
@@ -1,6 +1,6 @@
 package bio.fkaiser.fit3d.cli;
 
-import de.bioforscher.singa.structure.parser.pdb.rest.cluster.PDBSequenceCluster;
+import bio.singa.structure.parser.pdb.rest.cluster.PDBSequenceCluster;
 import org.apache.commons.cli.Option;
 import org.apache.commons.cli.OptionGroup;
 

diff --git a/fit3d-cli/src/main/resources/log4j.properties b/fit3d-cli/src/main/resources/log4j.properties
@@ -4,7 +4,7 @@ log4j.appender.stdout=org.apache.log4j.ConsoleAppender
 log4j.appender.stdout.Target=System.out
 log4j.appender.stdout.layout=org.apache.log4j.PatternLayout
 log4j.appender.stdout.layout.ConversionPattern=%d{HH:mm:ss} %5p [%c{1}:%M] - %m%n
-log4j.category.de.bioforscher.singa=ERROR
-log4j.category.de.bioforscher.singa.structure.parser.pdb.structures.StructureParser=WARN
+log4j.category.bio.singa=ERROR
+log4j.category.bio.singa.structure.parser.pdb.structures.StructureParser=WARN
 log4j.category.bio.fkaiser.mmm=ERROR
 log4j.category.bio.fkaiser.mmm.model.enrichment.IntraChainInteractionEnricher=WARN
diff --git a/fit3d-cli/src/test/java/bio/fkaiser/fit3d/cli/TemplateBasedCommandLineOptionsTest.java b/fit3d-cli/src/test/java/bio/fkaiser/fit3d/cli/TemplateBasedCommandLineOptionsTest.java
@@ -1,19 +1,16 @@
 package bio.fkaiser.fit3d.cli;
 
-import de.bioforscher.singa.core.utility.Resources;
+import bio.singa.core.utility.Resources;
 import org.junit.Before;
 import org.junit.Ignore;
 import org.junit.Rule;
 import org.junit.Test;
 import org.junit.rules.TemporaryFolder;
 
-import java.io.File;
 import java.io.IOException;
 import java.nio.file.Files;
 import java.nio.file.Paths;
 
-import static org.junit.Assert.assertTrue;
-
 /**
  * @author fk
  */
@@ -88,6 +85,22 @@ public void runWithMotifAndChainList() throws IOException {
         Fit3DCommandLine.main(commandLineArguments);
     }
 
+    @Test
+    public void runWithMotifAndChainListAndFiltering() throws IOException {
+
+        copyResource(MOTIF_RESOURCE);
+        copyResource(CHAIN_LIST_RESOURCE);
+
+        String[] commandLineArguments = new String[]{"template-based",
+                                                     "-m", folder.getRoot().toString() + "/" + MOTIF_RESOURCE,
+                                                     "-l", folder.getRoot().toString() + "/" + CHAIN_LIST_RESOURCE,
+                                                     "-r", "4.0",
+                                                     "-y", "3.0",
+                                                     "-f", folder.getRoot().toString() + "/summary.csv"};
+        Fit3DCommandLine.main(commandLineArguments);
+    }
+
+
     @Test
     public void runWithTargetFile() throws IOException {
 

diff --git a/fit3d-cli/src/test/java/bio/fkaiser/fit3d/cli/TemplateFreeCommandLineOptionsTest.java b/fit3d-cli/src/test/java/bio/fkaiser/fit3d/cli/TemplateFreeCommandLineOptionsTest.java
@@ -1,6 +1,6 @@
 package bio.fkaiser.fit3d.cli;
 
-import de.bioforscher.singa.core.utility.Resources;
+import bio.singa.core.utility.Resources;
 import org.junit.Ignore;
 import org.junit.Rule;
 import org.junit.Test;

diff --git a/fit3d-web/fit3d-web.iml b/fit3d-web/fit3d-web.iml
@@ -43,9 +43,9 @@
     <orderEntry type="library" name="Maven: log4j:log4j:1.2.17" level="project" />
     <orderEntry type="library" name="Maven: org.primefaces:primefaces:6.1" level="project" />
     <orderEntry type="library" name="Maven: org.primefaces.themes:bootstrap:1.0.10" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-structure:0.3.3" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-mathematics:0.3.3" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-features:0.3.3" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-structure:0.4.0" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-mathematics:0.4.0" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-features:0.4.0" level="project" />
     <orderEntry type="library" name="Maven: tec.uom:uom-se:1.0.8" level="project" />
     <orderEntry type="library" name="Maven: javax.measure:unit-api:1.0" level="project" />
     <orderEntry type="library" name="Maven: tec.uom.lib:uom-lib-common:1.0.2" level="project" />
@@ -59,7 +59,7 @@
     <orderEntry type="library" name="Maven: commons-lang:commons-lang:2.4" level="project" />
     <orderEntry type="library" scope="RUNTIME" name="Maven: org.apache.logging.log4j:log4j-api:2.6.2" level="project" />
     <orderEntry type="library" name="Maven: org.rcsb:mmtf-api:1.0.7" level="project" />
-    <orderEntry type="library" name="Maven: de.bioforscher.singa:singa-core:0.3.3" level="project" />
+    <orderEntry type="library" name="Maven: bio.singa:singa-core:0.4.0" level="project" />
     <orderEntry type="library" name="Maven: javax.mail:mail:1.4.7" level="project" />
     <orderEntry type="library" name="Maven: javax.activation:activation:1.1" level="project" />
     <orderEntry type="library" name="Maven: org.omnifaces:omnifaces:1.10" level="project" />

diff --git a/fit3d-web/pom.xml b/fit3d-web/pom.xml
@@ -4,9 +4,9 @@
          xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
     <modelVersion>4.0.0</modelVersion>
 
-    <groupId>de.bioforscher.fit3d</groupId>
+    <groupId>bio.fkaiser.fit3d</groupId>
     <artifactId>fit3d-web</artifactId>
-    <version>0.1.3</version>
+    <version>0.2.0</version>
     <packaging>war</packaging>
 
     <properties>
@@ -68,14 +68,14 @@
             <version>1.0.10</version>
         </dependency>
         <dependency>
-            <groupId>de.bioforscher.singa</groupId>
+            <groupId>bio.singa</groupId>
             <artifactId>singa-structure</artifactId>
-            <version>0.3.3</version>
+            <version>0.4.0</version>
         </dependency>
         <dependency>
-            <groupId>de.bioforscher.singa</groupId>
+            <groupId>bio.singa</groupId>
             <artifactId>singa-core</artifactId>
-            <version>0.3.3</version>
+            <version>0.4.0</version>
         </dependency>
         <dependency>
             <groupId>javax.mail</groupId>