bedtools/bedtools_sample

genomics

sampling

subset

reservoir

random

quality-control

Description

Take a random sample of records from BED/GFF/VCF/BAM files using reservoir sampling algorithm.

bedtools sample uses the reservoir sampling algorithm to randomly select a specified number
of records from genomic interval files. This is particularly useful for creating representative
subsets of large datasets for testing, quality control, or downstream analysis.

This tool is commonly used for:

Creating representative subsets of large genomic datasets
Quality control and validation with smaller sample sizes
Testing pipelines with manageable data volumes
Generating training datasets for machine learning applications
Reducing file sizes while maintaining statistical properties

Type

bash_script

License

MIT

Keywords

genomics

sampling

subset

reservoir

random

quality-control

Links

Repository

Documentation

Contributors

Launch

Run this component

Run the following command to execute this component with Nextflow:

cat > params.yaml <<'EOM'  
output: "$id.$key.output.bed"  
id: "run"  
publish_dir: "output/"  
EOM

nextflow run https://packages.viash-hub.com/vsh/biobox.git \  
  -revision v0.4.0 \  
  -main-script target/nextflow/bedtools/bedtools_sample/main.nf \  
  -params-file params.yaml

Inputs

Name	Type & Properties	Description
--input -i	file required	Input file to sample from. Format: BED, GFF, VCF, or BAM file Content: Genomic intervals or alignments to sample from Usage: Records will be randomly selected from this file Requirements: File must be in valid format for the specified type

Name

Type & Properties

--input

-i

file

required

Outputs

Name	Type & Properties	Description
--output	file required output	Output file containing the sampled records. Format: Same format as input file Content: Randomly selected subset of input records Size: Contains exactly n records (unless input has fewer records)

Name

Type & Properties

--output

file

required

output

Sampling Options

Name	Type & Properties	Description
--number -n	integer	Number of records to randomly sample. Default: 1,000,000 Range: 1 to total number of records in input Usage: If input has fewer records than requested, all records are returned Memory: All selected records held in memory before output
--seed	integer	Integer seed for random number generation. Usage: Ensures reproducible random sampling Range: Any integer value Default: Automatically chosen seed (non-reproducible) Applications: Reproducible research, testing, validation

Name

Type & Properties

--number

-n

integer

--seed

integer

Strand Options

Name	Type & Properties	Description
--strand_requirement -s	string	Require records from specific strand orientation. Values: "forward", "reverse", or unspecified for both strands Usage: Only sample records from the specified strand Requirements: Input must contain strand information Applications: Strand-specific analyses, RNA-seq processing

Name

Type & Properties

--strand_requirement

-s

string

Output Format Options

Name	Type & Properties	Description
--output_bed -bed	boolean_true	Convert BAM input to BED format output. Usage: Only applicable when input is BAM format Effect: Output genomic coordinates in BED format instead of BAM Applications: Converting alignment data to interval format Default: false (maintain input format)
--uncompressed_bam -ubam	boolean_true	Write uncompressed BAM output. Usage: Only applicable when input is BAM format Effect: Output BAM file without compression Trade-off: Faster writing but larger file size Default: false (compressed BAM output)
--include_header -header	boolean_true	Include the original file header in output. Usage: Preserves metadata from input file Applications: Maintaining file structure and annotations Formats: Particularly relevant for VCF and GFF files Default: false (no header included)

Name

Type & Properties

--output_bed

-bed

boolean_true

--uncompressed_bam

-ubam

boolean_true

--include_header

-header

boolean_true

Performance Options

Name	Type & Properties	Description
--no_buffer -nobuf	boolean_true	Disable output buffering for real-time processing. Effect: Each line printed immediately instead of buffered Trade-off: Slower output but enables real-time processing Applications: Pipeline integration, streaming processing Default: false (buffered output for performance)
--input_buffer -iobuf	string	Amount of memory to allocate for input buffer. Format: Integer with optional K/M/G suffix Examples: "1G", "512M", "2048K" Usage: Larger buffers can improve I/O performance Note: Currently has no effect with compressed files

Name

Type & Properties

--no_buffer

-nobuf

boolean_true

--input_buffer

-iobuf

string