SnpSift Intervals

Filter VCF entries that intersect (or don't intersect) genomic intervals defined in BED files.

Usage

java -jar SnpSift.jar intervals [options] file_1.bed [file_2.bed ... file_N.bed]
Options:
    -i <file.vcf> : VCF input file. Default: STDIN
    -x            : Exclude mode: output VCF entries that do NOT intersect any interval

Warning

The VCF input must be provided via -i or STDIN. All positional arguments are treated as BED files. If you pass a VCF file as a positional argument, it will be incorrectly loaded as a BED file.

How it works

All intervals from all BED files are loaded into a single interval forest. Each VCF entry is tested for position overlap against this interval set.

By default, VCF entries that overlap any interval are output. With -x, VCF entries that do NOT overlap any interval are output. Output is in standard VCF format (with header).

Example

Keep only variants in target regions:

cat variants.vcf | java -jar SnpSift.jar intervals target_regions.bed > variants_on_target.vcf

Exclude variants in problematic regions:

java -jar SnpSift.jar intervals -x -i variants.vcf blacklist.bed > variants_filtered.vcf

Multiple BED files can be provided and their intervals are combined:

java -jar SnpSift.jar intervals -i variants.vcf peaks_rep1.bed peaks_rep2.bed peaks_rep3.bed > variants_in_peaks.vcf

BED file format

BED files are tab-separated with 0-based coordinates (end position is exclusive). Only the first three columns are used (chr, start, end); additional columns are ignored for filtering purposes. A minimum of two columns (chr, start) is required; if end is missing, it defaults to start. Header lines must start with #.