intrepiditee / forwardsimulatorhjlib

A Forward Simulator for IBD Sharing

1. Simulation

1.Usage

bash run.sh --simulate numberOfGenerations firstGenerationToStore lastGenerationToStore generationSize numberOfThreads

This will simulate for "numOfGeneration" generations including the founder generation. Each generation has "generationSize" individuals, of which half are males and the other half are females. It will store the IBD segments, the genomes, mutated sites, and parent IDs of individuals from Generation "firstGenerationToStore" to Generation "lastGenerationToStore" (starting from zero). Genomes and mutated sites are stored in binary for further processing. "numberOfThreads" is the maximum number of threads that will be spawned during the execution.

Example:

bash run.sh --simulate 10 6 9 1000 22

This will simulate for 10 generations each of 1000 individuals using 22 threads. The last 4 generations (6, 7, 8, 9) will be stored.

2. File Output

All the genomes, mutations sites, and parent IDs will be generated in ForwardSimulatorHjLib/target/out in binary. These files are required for computing pairwise distances and generating VCF files. They are used internally and are not useful to the users.

All the IBD segments will be generated in ForwardSimulatorHjLib/target/ibd in text. Each file is tab delimited. The first line of each file is a header.

2. Subsetting UK Biobank VCF Files

This must be done before parsing founder sequences.

Alternatively, you can use VCFtools to do this. Just make sure input files comply to the requirements described in Step 3.

1.Usage

Example:

2. File Output

3. Parsing Founder Sequences

This must be done before generating VCF files from the simulation.

The input files must be in VCF format. There must be 22 files corresponding to the 22 chromosomes. The filenames must be in the form of chr{}.recode.vcf, where {} needs to be replaced by the chromosome number. The input files must be supplied under ForwardSimulatorHjLib/target/subset/.

The input VCF files can contain different number of sites. The number of sites contained in one input VCF file for one chromosome equals the number sites that will be contained in the VCF file for that chromosome generated from the simulation. Note that this step does not generate VCF files from the simulation yet. This step extracts necessary information from the founder generation to prepare for generating VCF files from the simulation.

The number of individuals contained in each input VCF file must be same as the input to the simulation.

1.Usage

bash run.sh --parse generationSize numberOfThread

Example:

bash run.sh --parse 1000 22

2. File Output

All the sites and bases in the input files will be stored under ForwardSimulatorHjLib/target/ukb/ in binary. They are used internally and are not useful to the users.

4. Generating VCF Files from Simulation

1.Usage

bash run.sh --generate generationSize firstGenerationToStore lastGenerationToStore numberOfThreads

"firstGenerationToStore" and "lastGenerationToStore" must match the inputs to the simulation.

Example:

bash run.sh --generate 1000 6 9 22

This will generate the VCF files for the last 4 generations (6, 7, 8, 9) stored during the simulation, using 22 threads.

2. File Output

The output VCF files will be stored under ForwardSimulatorHjLib/target/final. They can be used as inputs to RaPID.

5. Generating Mapping Files for RaPID

1.Usage

bash run.sh --map

2. File Output

The mapping files (from chromosome positions to genetic positions) will be generated under ForwardSimulatorHjLib/target/map/. You can use these files as input to RaPID.

6. Computing Pairwise Distances

1.Usage

Example:

2. File Output