Add geneBlacklistFilter, SPLICE_AI and ALPHA_MISSENSE to example YAML…

… files.

Return exit code from exomiser-cli

exomiser-cli/CHANGELOG.md

@@ -2,7 +2,10 @@
 
 ## 14.0.0 2023-MM-dd
 
+- Minimum Java version is now set to __Java 17__
 - Enabled independent update of ClinVar data [#501](https://github.com/exomiser/Exomiser/issues/501)
+- Fix for issue [#531](https://github.com/exomiser/Exomiser/issues/531) where the `priorityScoreFilter` and `regulatoryFeatureFilter` pass/fail counts were not displayed in the HTML.
+- Fix for issue [#534](https://github.com/exomiser/Exomiser/issues/534) where variant frequency and/or pathogenicity annotations are missing in certain run configurations.
 
 
 ## 13.3.0 2023-10-17

exomiser-cli/src/main/java/org/monarchinitiative/exomiser/cli/BatchFileReader.java

@@ -33,8 +33,6 @@
 import java.util.function.Predicate;
 import java.util.stream.Stream;
 
-import static java.util.stream.Collectors.toList;
-
 /**
  * Reads in Exomiser batch files and returns a list of Paths to the
  * settings/analysis files. The reader expects a single path per line.
@@ -56,7 +54,7 @@ public static List<Path> readPathsFromBatchFile(Path batchFile) {
                     .filter(commentLines())
                     .filter(emptyLines())
                     .map(line -> Paths.get(line.trim()))
-                    .collect(toList());
+                    .toList();
         } catch (IOException ex) {
             logger.error("Unable to read batch file {}", batchFile, ex);
         }
@@ -71,7 +69,7 @@ public static List<JobProto.Job> readJobsFromBatchFile(Path batchFile) {
                     .filter(commentLines())
                     .filter(emptyLines())
                     .flatMap(line -> commandLineJobReader.readJobs(CommandLineOptionsParser.parse(line.split("\\s+"))).stream())
-                    .collect(toList());
+                    .toList();
         } catch (IOException ex) {
             logger.error("Unable to read batch file {}", batchFile, ex);
         }

exomiser-cli/src/main/java/org/monarchinitiative/exomiser/cli/CommandLineJobReader.java

@@ -70,7 +70,7 @@ public List<JobProto.Job> readJobs(CommandLine commandLine) {
         if (userOptions.equals(Set.of("analysis-batch"))) {
             Path analysisBatchFile = Path.of(commandLine.getOptionValue("analysis-batch"));
             List<Path> analysisScripts = BatchFileReader.readPathsFromBatchFile(analysisBatchFile);
-            return analysisScripts.stream().map(JobReader::readJob).collect(Collectors.toList());
+            return analysisScripts.stream().map(JobReader::readJob).toList();
         }
         // new batch option which will parse each line as a cli command
         if (userOptions.equals(Set.of("batch"))) {
@@ -361,16 +361,12 @@ private <U extends Message.Builder> U tryParseJsonOrYaml(U messageBuilder, Path
     }
 
     private AnalysisProto.Preset parsePreset(String presetValue) {
-        switch (presetValue.toLowerCase()) {
-            case "exome":
-                return AnalysisProto.Preset.EXOME;
-            case "genome":
-                return AnalysisProto.Preset.GENOME;
-            case "phenotype-only":
-                return AnalysisProto.Preset.PHENOTYPE_ONLY;
-            default:
-                throw new IllegalArgumentException("Unrecognised preset option: " + presetValue);
-        }
+        return switch (presetValue.toLowerCase()) {
+            case "exome" -> AnalysisProto.Preset.EXOME;
+            case "genome" -> AnalysisProto.Preset.GENOME;
+            case "phenotype-only" -> AnalysisProto.Preset.PHENOTYPE_ONLY;
+            default -> throw new IllegalArgumentException("Unrecognised preset option: " + presetValue);
+        };
     }
 
     private OutputProto.OutputOptions readOutputOptions(Path outputOptionsPath) {

exomiser-cli/src/main/java/org/monarchinitiative/exomiser/cli/CommandLineOptionsParser.java

@@ -28,8 +28,6 @@
 import java.nio.file.Paths;
 import java.util.List;
 
-import static java.util.stream.Collectors.toList;
-
 /**
  * @author Jules Jacobsen <[email protected]>
  * @since 13.0.0
@@ -226,7 +224,7 @@ protected static List<String> fileDependentOptions() {
         return options.getOptions().stream()
                 .filter(option -> "file".equals(option.getArgName()))
                 .map(Option::getLongOpt)
-                .collect(toList());
+                .toList();
     }
 
     public static void printHelp() {

exomiser-cli/src/main/java/org/monarchinitiative/exomiser/cli/Main.java

@@ -50,9 +50,9 @@ public static void main(String[] args) {
 
         // all ok so far - try launching the app
         Locale.setDefault(Locale.UK);
-        SpringApplication.run(Main.class, args).close();
-
+        int exitCode = SpringApplication.exit(SpringApplication.run(Main.class, args));
         logger.info("Exomising finished - Bye!");
+        System.exit(exitCode);
     }
 
 }

exomiser-cli/src/main/resources/application.properties

@@ -62,7 +62,7 @@ exomiser.hg19.data-version=2402
 #exomiser.hg38.use-clinvar-white-list=true
 
 ### phenotypes ###
-exomiser.phenotype.data-version=2109
+exomiser.phenotype.data-version=2402
 #exomiser.phenotype.data-directory=${exomiser.data-directory}/${exomiser.phenotype.data-version}_phenotype
 # String random walk data file
 #exomiser.phenotype.random-walk-file-name=rw_string_10.mv

exomiser-cli/src/main/resources/examples/preset-exome-analysis.yml

@@ -33,11 +33,11 @@ frequencySources: [
   #  GNOMAD_G_OTH,
     GNOMAD_G_SAS
 ]
-# Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM
+# Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM, SPLICE_AI, ALPHA_MISSENSE
 # REMM is trained on non-coding regulatory regions
 # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
 # and updated their location in the application.properties. Exomiser will not run without this.
-pathogenicitySources: [ REVEL, MVP ]
+pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
 #this is the standard exomiser order.
 steps: [
     failedVariantFilter: { },

exomiser-cli/src/main/resources/examples/preset-genome-analysis.yml

@@ -33,11 +33,11 @@ frequencySources: [
   #  GNOMAD_G_OTH,
     GNOMAD_G_SAS
 ]
-# Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM
+# Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM, SPLICE_AI, ALPHA_MISSENSE
 # REMM is trained on non-coding regulatory regions
 # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
 # and updated their location in the application.properties. Exomiser will not run without this.
-pathogenicitySources: [ REVEL, MVP, REMM ]
+pathogenicitySources: [ REVEL, MVP, REMM, SPLICE_AI ]
 # this is the recommended order for a genome-sized analysis.
 steps: [
     hiPhivePrioritiser: { },

exomiser-cli/src/main/resources/examples/test-analysis-exome.yml

@@ -51,11 +51,11 @@ analysis:
       #  GNOMAD_G_OTH,
         GNOMAD_G_SAS
     ]
-  # Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM
+  # Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM, SPLICE_AI, ALPHA_MISSENSE
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-    pathogenicitySources: [ REVEL, MVP ]
+    pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
   # this is the standard exomiser order.
   # all steps are optional
     steps: [
@@ -65,6 +65,7 @@ analysis:
       # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
       #intervalFilter: {bed: /full/path/to/bed_file.bed},
       #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
+      # geneBlacklistFilter: { },
         failedVariantFilter: { },
       #qualityFilter: {minQuality: 50.0},
         variantEffectFilter: {

exomiser-cli/src/main/resources/examples/test-analysis-genome.yml

@@ -51,11 +51,11 @@ analysis:
       #  GNOMAD_G_OTH,
         GNOMAD_G_SAS
     ]
-  # Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM
+  # Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM, SPLICE_AI, ALPHA_MISSENSE
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-    pathogenicitySources: [ REVEL, MVP, REMM ]
+    pathogenicitySources: [ REVEL, MVP, REMM, SPLICE_AI ]
   # this is the recommended order for a genome-sized analysis.
   # all steps are optional
     steps: [
@@ -66,6 +66,7 @@ analysis:
       # 0.501 is a good compromise to select good phenotype matches and the best protein-protein interactions hits from hiPhive
         priorityScoreFilter: { priorityType: HIPHIVE_PRIORITY, minPriorityScore: 0.501 },
         failedVariantFilter: { },
+      # geneBlacklistFilter: { },
       #intervalFilter: {interval: 'chr10:123256200-123256300'},
       # or for multiple intervals:
       #intervalFilter: {intervals: ['chr10:123256200-123256300', 'chr10:123256290-123256350']},

exomiser-cli/src/main/resources/examples/test-analysis-multisample.yml

@@ -50,11 +50,11 @@ analysis:
       #  GNOMAD_G_OTH,
         GNOMAD_G_SAS
     ]
-  # Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM
+  # Possible pathogenicitySources: (POLYPHEN, MUTATION_TASTER, SIFT), (REVEL, MVP), CADD, REMM, SPLICE_AI, ALPHA_MISSENSE
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-    pathogenicitySources: [ REVEL, MVP ]
+    pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
   # this is the standard exomiser order.
   # all steps are optional
     steps: [
@@ -64,6 +64,7 @@ analysis:
       # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
       #intervalFilter: {bed: /full/path/to/bed_file.bed},
       #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
+      # geneBlacklistFilter: { },
         failedVariantFilter: { },
       #qualityFilter: {minQuality: 50.0},
         variantEffectFilter: {

exomiser-cli/src/test/java/org/monarchinitiative/exomiser/cli/CommandLineJobReaderTest.java

@@ -52,8 +52,7 @@
 import static org.hamcrest.Matchers.equalTo;
 import static org.hamcrest.Matchers.containsInAnyOrder;
 import static org.junit.jupiter.api.Assertions.assertThrows;
-import static org.monarchinitiative.exomiser.core.model.pathogenicity.PathogenicitySource.MVP;
-import static org.monarchinitiative.exomiser.core.model.pathogenicity.PathogenicitySource.REVEL;
+import static org.monarchinitiative.exomiser.core.model.pathogenicity.PathogenicitySource.*;
 
 /**
  * @author Jules Jacobsen <[email protected]>
@@ -265,7 +264,7 @@ class CommandLineJobReaderTest {
 //                    FrequencySource.GNOMAD_G_OTH,
                     FrequencySource.GNOMAD_G_SAS
             ))
-            .pathogenicitySources(ImmutableSet.of(REVEL, MVP))
+            .pathogenicitySources(ImmutableSet.of(REVEL, MVP, SPLICE_AI))
             .addFailedVariantFilter()
             .addVariantEffectFilter(ImmutableSet.of(
                     FIVE_PRIME_UTR_EXON_VARIANT,

exomiser-cli/src/test/resources/exome-analysis.yml

@@ -42,7 +42,7 @@ frequencySources: [
 # REMM is trained on non-coding regulatory regions
 # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
 # and updated their location in the application.properties. Exomiser will not run without this.
-pathogenicitySources: [ REVEL, MVP ]
+pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
 #this is the standard exomiser order.
 #all steps are optional
 steps: [
@@ -54,7 +54,8 @@ steps: [
   # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
   #intervalFilter: {bed: /full/path/to/bed_file.bed},
   #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
-    failedVariantFilter: { },
+  # geneBlacklistFilter: { },
+  failedVariantFilter: { },
   #qualityFilter: {minQuality: 50.0},
     variantEffectFilter: {
       remove: [

exomiser-cli/src/test/resources/pfeiffer-analysis-v8-12.yml

@@ -54,7 +54,7 @@ analysis:
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-  pathogenicitySources: [ REVEL, MVP ]
+  pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
   #this is the standard exomiser order.
   #all steps are optional
   steps: [
@@ -66,6 +66,7 @@ analysis:
     # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
     #intervalFilter: {bed: /full/path/to/bed_file.bed},
     #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
+    # geneBlacklistFilter: { },
       failedVariantFilter: { },
     #qualityFilter: {minQuality: 50.0},
       variantEffectFilter: {

exomiser-cli/src/test/resources/pfeiffer-job-phenopacket.yml

@@ -93,7 +93,7 @@ analysis:
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-  pathogenicitySources: [ REVEL, MVP ]
+  pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
   #this is the standard exomiser order.
   #all steps are optional
   steps: [
@@ -105,6 +105,7 @@ analysis:
     # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
     #intervalFilter: {bed: /full/path/to/bed_file.bed},
     #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
+    # geneBlacklistFilter: { },
       failedVariantFilter: { },
     #qualityFilter: {minQuality: 50.0},
       variantEffectFilter: {

exomiser-cli/src/test/resources/pfeiffer-job-sample.yml

@@ -61,7 +61,7 @@ analysis:
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-  pathogenicitySources: [ REVEL, MVP ]
+  pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
   #this is the standard exomiser order.
   #all steps are optional
   steps: [
@@ -73,6 +73,7 @@ analysis:
     # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
     #intervalFilter: {bed: /full/path/to/bed_file.bed},
     #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
+    # geneBlacklistFilter: { },
       failedVariantFilter: { },
     #qualityFilter: {minQuality: 50.0},
       variantEffectFilter: {

exomiser-cli/src/test/resources/test-analysis-exome.yml

@@ -55,7 +55,7 @@ analysis:
   # REMM is trained on non-coding regulatory regions
   # *WARNING* if you enable CADD or REMM ensure that you have downloaded and installed the CADD/REMM tabix files
   # and updated their location in the application.properties. Exomiser will not run without this.
-  pathogenicitySources: [ REVEL, MVP ]
+  pathogenicitySources: [ REVEL, MVP, SPLICE_AI ]
   # this is the standard exomiser order.
   # all steps are optional
   steps: [
@@ -65,6 +65,7 @@ analysis:
     # or using a BED file - NOTE this should be 0-based, Exomiser otherwise uses 1-based coordinates in line with VCF
     #intervalFilter: {bed: /full/path/to/bed_file.bed},
     #genePanelFilter: {geneSymbols: ['FGFR1','FGFR2']},
+    # geneBlacklistFilter: { },
     failedVariantFilter: { },
     #qualityFilter: {minQuality: 50.0},
     variantEffectFilter: {