The mutation accumulation database
The mutation accumulation database
mutacc is a tool that makes it possible to create synthetic datasets to be used for quality control or benchmarking of bioinformatic tools and pipelines intended for variant calling of clinical variants. Using raw reads that supports a known variant from a real NGS data, mutacc stores the relevant reads from each case into a database. This database can then be queried to create synthetic datasets that can be used as positive controls bioinformatics pipelines.
Running the app using Docker (No installation of any software or database required)
An example containing a demo setup for the app is included in the docker-compose file. Note that this file is not intended for use in production and is only provided to illustrate how an image containing the application could be connected to a MongoDB instance and perform commands provided when running it as a container. A Docker image file for Mutacc can be pulled from Docker Hub, or can be built from the Dockerfile provided in the GitHub repository folder. Start the docker-compose demo using this command:
docker-compose up -d
What the docker-compose command does:
- Starts the database
- extracts the reads from a demo case (demo case resources are located under /mutacc/resources)
- Saves them to database
- Exports them from the database to a local file
When the above command is executed, it creates the following 4 directories:
variants in the working directory. The directory names
variants contains the vcf with the variants of interest for this demo case.
After running the test, don't forget to run docker-compose to remove containers, networks, volumes and images created by docker-compose.
For installation of mutacc and the external prerequisites, this is made easy by creating conda environment
conda create -n <env_name> python=3.6 pip numpy cython
source activate <env_name>
conda install -c bioconda picard conda install -c bioconda seqkit
Within the conda environment, do
pip install mutacc
To install from PyPI, or clone this repo and install
pip install git+https://github.com/Clinical-Genomics/mutacc
Some options are best passed to mutacc through a configuration file. below is an example of a config file, using the YAML format.
#EXAMPLE OF A CONFIGURATION FILE host: <host> #Defaults to 'localhost' port: <port> #Defaults to 27017 database: <database_name> #Defaults to 'mutacc' username: <username> password: <password> root_dir: <path_to_root>
The 'root_dir' entry specifies an existing directory in the file system, where all files generated by mutacc will be stored in corresponding subdirectories. E.g. all generated fastq files will be stored in /.../root_dir/reads/
Populate the mutacc database
To export data sets from the mutacc DB, the database must first be populated. To extract the raw reads supporting a known variant, mutacc takes use some relevant files generated from a NGS experiment up to the variant calling itself. That is the bam file, and vcf file containing only the variants of interest.
This information is specified as a 'case', represented in yaml format
#EXAMPLE OF A CASE #THE CASE FIELD CONTAINS METADATA OF THE CASE ITSELF case: case_id: 'case123' #REQUIRED CASE_ID #LIST OF THE SAMPLES INVOLVED IN THE EXPERIMENT (MAY BE ONE, OR SEVERAL, E.G. #A TRIO) samples: - sample_id: 'sample1' #REQUIRED analysis_type: 'wgs' #REQUIRED sex: 'male' #REQUIRED mother: 'sample2' #REQUIRED (CAN BE 0 if no mother) father: 'sample3' #REQUIRED (CAN BE 0 if no father) bam_file: /path/to/sorted_bam #REQUIRED phenotype: 'affected' - sample_id: 'sample2' analysis_type: 'wgs' sex: 'female' mother: '0' #0 if no parent father: '0' bam_file: /path/to/sorted_bam phenotype: 'unaffected' - sample_id: 'sample2' analysis_type: 'wgs' sex: 'male' mother: '0' father: '0' bam_file: /path/to/sorted_bam phenotype: 'affected' #PATH TO VCF FILE CONTAINING VARIANTS OF INTEREST FROM CASE variants: /path/to/vcf
This will find the reads from the bam files specified for each sample. If it is desired that the reads are found from the fastq files instead, this can be done by specifying the fastq-files as such
- sample_id: 'sample1' analysis_type: 'wgs' sex: 'male' mother: 'sample2' father: 'sample3' bam_file: /path/to/sorted_bam fastq_files: - /path/to/fastq1 - /path/to/fastq2 phenotype: 'affected'
To extract the reads from the case
mutacc --config-file <config_file> extract --padding 600 --case <case_file>
the --padding option takes the number of basepairs that the desired region is padded with.
This will create a file <case_id>.json stored in the directory specified in the /.../root_dir/imports directory.
To import the case into the database
mutacc db import /.../root_dir/imports/<case_id>.json
The db command is called each time mutacc needs to do any operation against the database.
This will try to establish a connection to an instance of mongodb, by default running on 'localhost' on port 27017. If this is not wanted, it can be specified with the --host and --port options.
mutacc db -h <host> -p <port> import <case_id>.json
If authentication is required, this can be specified with the --username and --password options.
or in a configuration file e.g.
host: <host> port: <port> username: <username> password: <password>
mutacc --config-file <config.yaml> db import <case_id>.json
Export datasets from the database
The datasets are exported one sample at the time. To export a synthetic dataset, the export command is used together with options.
Usage: mutacc db export [OPTIONS] exports dataset from DB Options: -c, --case-mongo TEXT mongodb query language json-string to query for cases in database -v, --variant-mongo TEXT mongodb query language json-string to query for variants in database -t, --variant-type TEXT Type of variant -a, --analysis [wes|wgs] Type of analysis --all-variants Export all variants in database -m, --member [father|mother|child|affected] Type of sample -s, --sex [male|female] Sex of sample --vcf-dir PATH Directory where vcf is created. Defaults to mutacc-root/variants -p, --proband Variants from all affected samples, regardless of pedigree -n, --sample-name TEXT Name of sample -j, --json-out Print results to stdout as json-string --help Show this message and exit.
mutacc --config-file <config.yaml> db export -m affected --all-variants
will find all the cases from the mutacc DB, and store this information in a file /.../root_dir/queries/sample_name_query.mutacc.
to export an entire trio, this can be done by
mutacc --config-file <config_file> db export -m child --all-variants -p -n child mutacc --config-file <config_file> db export -m father --all-variants -n father mutacc --config-file <config_file> db export -m mother --all-variants -n mother
This will create three files child_query_mutacc.json, father_query_mutacc.json, and mother_query_mutacc.json.
the export subcommand will also generate a truth set vcf-file for each exported samples, containing all queried variants.
To make a dataset (fastq-files) from a query file the synthesize command is used with the following options
Path to the bam file for sample to be used as background
Path to fastq file for sample to be used as background
Path to second fastq file (if paired end experiment)
Path to the query json-files created with the export command
Directory where fastq files will be stored. defaults to /.../root_dir/datasets
example, using the query files created above
mutacc --config-file <config_file> synthesize -b <bam> -f <fastq1_child> -f2 <fastq2_child> -q child_query_mutacc.json mutacc --config-file <config_file> synthesize -b <bam> -f <fastq1_father> -f2 <fastq2_father> -q father_query_mutacc.json mutacc --config-file <config_file> synthesize -b <bam> -f <fastq1_mother> -f2 <fastq2_mother> -q mother_query_mutacc.json
The created fastq-files will be stored in the directory /.../root_dir/datasets/ or in directory specified by ---dataset-dir
Remove case from database
To remove a case from the mutacc DB, and all the generated bam, and fastq files generated from that case from disk, the remove command is used
mutacc --config-file <config.yaml> db remove <case_id>
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.