Master: Development:

The Exomiser is a Java program that finds potential disease-causing variants from whole-exome or whole-genome sequencing data.

Starting from a VCF file and a set of phenotypes encoded using the Human Phenotype Ontology (HPO) it will annotate, filter and prioritise likely causative variants. The program does this based on user-defined criteria such as a variant's predicted pathogenicity, frequency of occurrence in a population and also how closely the given phenotype matches the known phenotype of diseased genes from human and model organism data.

The functional annotation of variants is handled by Jannovar and uses UCSC KnownGene transcript definitions and hg19 genomic coordinates.

Variants are prioritised according to user-defined criteria on variant frequency, pathogenicity, quality, inheritance pattern, and model organism phenotype data. Predicted pathogenicity data is extracted from the dbNSFP resource. Variant frequency data is taken from the 1000 Genomes, ESP and ExAC datasets. Subsets of these frequency and pathogenicity data can be defined to further tune the analysis. Cross-species phenotype comparisons come from our PhenoDigm tool powered by the OWLTools OWLSim algorithm.

The Exomiser was developed by the Computational Biology and Bioinformatics group at the Institute for Medical Genetics and Human Genetics of the Charité - Universitätsmedizin Berlin, the Mouse Informatics Group at the Sanger Institute and other members of the Monarch initiative.

The prebuilt Exomiser binaries can be obtained from the releases page and supporting data files can be downloaded from the Exomiser FTP site.

It is possible to use the same data sources for each major version, in order to avoid having to download the data files for each software point release. To do this, edit the exomiser.data-directory field in the application.properties file to point to the data directory of the other installation.

For example you have an exomiser installation located at /opt/exomiser-cli-7.0.0 which contains the data files in the directory /opt/exomiser-cli-7.0.0/data. You can use the release 7.2.3 (same major version) by unzipping the release to /opt/exomiser-cli-7.2.3 and changing the line in the file /opt/exomiser-cli-7.2.3/application.properties from

#root path where data is to be downloaded and worked on
#it is assumed that all the files required by exomiser listed in this properties file
#will be found in the data directory unless specifically overridden here.
exomiser.data-directory=data

to

exomiser.data-directory=/opt/exomiser-cli-7.0.0/data

For further instructions on installing and running please refer to the README.md file.

Please refer to the manual for details on how to configure and run the Exomiser.

There is a limited demo version of the exomiser hosted by the Monarch Initiative. This instance is for teaching purposes only and is limited to small exome analysis.

The exomiser can also be used as a library in Spring Java applications. Add the exomiser-spring-boot-starter library to your pom/gradle build script.

In your configuration class add the @EnableExomiser annotation

@EnableExomiser
public class MainConfig {
   
}

Or if using Spring boot for your application, you can add it on your main class

@EnableExomiser
@SpringBootApplication
public class Application {
    public static void main(String[] args) {
        SpringApplication.run(Application.class, args);
    }
}

In your application use the AnalysisBuilder obtained from the Exomiser instance to configure your analysis. Then run the Analysis using the Exomiser class. Creation of the Exomiser is a complicated process so defer this to Spring and the exomiser-spring-boot-starter. Calling the add prefixed methods will add that analysis step to the analysis in the order that they have been defined in your code.

Example usage:

@Autowired
private final Exomiser exomiser;

...
           
    Analysis analysis = exomiser.getAnalysisBuilder()
                .vcfPath(vcfPath)
                .pedPath(pedPath)
                .probandSampleName(probandSampleId)
                .hpoIds(phenotypes)
                .analysisMode(AnalysisMode.PASS_ONLY)
                .modeOfInheritance(ModeOfInheritance.AUTOSOMAL_DOMINANT)
                .frequencySources(FrequencySource.ALL_EXTERNAL_FREQ_SOURCES)
                .pathogenicitySources(EnumSet.of(PathogenicitySource.POLYPHEN, PathogenicitySource.MUTATION_TASTER, PathogenicitySource.SIFT))
                .addPhivePrioritiser()
                .addPriorityScoreFilter(PriorityType.PHIVE_PRIORITY, 0.501f)
                .addQualityFilter(500.0)
                .addRegulatoryFeatureFilter()
                .addFrequencyFilter(0.01f)
                .addPathogenicityFilter(true)
                .addInheritanceFilter()
                .addOmimPrioritiser()
                .build();
                
    AnalysisResults analysisResults = exomiser.run(analysis);

Analysing whole genomes using the AnalysisMode.FULL or AnalysisMode.SPARSE will use a lot of RAM (~16GB for 4.5 million variants without any extra variant data being loaded) the standard Java GC will fail to cope well with these. Using the G1GC should solve this issue. e.g. add -XX:+UseG1GC to your java -jar -Xmx... incantation.