ShredPacBio
Shreds long single-molecule reads (e.g. PacBio HiFi) into fake paired-end reads, walking each molecule front-to-back and using every base exactly once.
Basic Usage
shredpacbio.sh in=<file> out=<pairs> outs=<singletons> outd=<discards>Converts long single-ended reads into fake paired-end reads for use with paired-end-oriented tools such as CallVariants. Each 2×readlen window of a molecule becomes one pair: the first readlen bases form read 1, the next readlen bases form read 2 (reverse-complemented by default so the pair is FR-oriented, matching real Illumina data). This lets a long span be placed with two short, cheap alignments instead of one long one, which helps in repetitive regions.
Parameters
Parameters control input/output files and the read-length/residual-handling behavior of the shredding process.
File Parameters
- in=<file>
- Input long reads (FASTQ), single-ended. Required parameter.
- out=<file>
- Interleaved proper pairs (FASTQ). Required parameter.
- outs=<file>
- Singletons: a leftover read from a molecule with no mate >= minlen. Optional.
- outd=<file>
- Discards: leftover fragments shorter than minlen. Optional.
Processing Parameters
- readlen=500
- Length of each emitted read; a full pair consumes 2×readlen bases from the source molecule. Aliases: rl, length, len.
- minlen=200
- Minimum length for any emitted read; shorter residuals are sent to discards. Must be <= readlen. Aliases: ml.
- rcompmate=t
- Reverse-complement read 2 (and reverse its quality) so each pair is FR-oriented like real Illumina data. Aliases: rcomp, rc.
- reads=-1
- If nonnegative, stop after this many input molecules.
- overwrite=t
- Permit overwriting existing output files. Alias: ow.
Examples
Basic Pairing
shredpacbio.sh in=pacbio.fastq out=pairs.fastqShreds PacBio HiFi reads into 500bp fake pairs using default settings; residuals below 200bp are silently dropped since no outs/outd is given.
Capturing Singletons and Discards
shredpacbio.sh in=hifi_reads.fastq out=pairs.fastq outs=singles.fastq outd=discards.fastq readlen=250Uses a shorter 250bp read length and keeps every leftover fragment, routing unmatched reads to singles.fastq and sub-minlen fragments to discards.fastq.
Disabling Mate Reverse-Complement
shredpacbio.sh in=reads.fastq out=pairs.fastq rcompmate=fEmits both reads of each pair in their original forward orientation instead of the default FR orientation.
How It Works
Each molecule is walked front-to-back in fixed 2×readlen windows; every base is used exactly once, so there is no coverage loss or duplication the way overlapping-shred approaches have. What happens to the residual left after the last full window depends on its length:
- Residual < minlen: sent to discards (outd).
- minlen <= residual < readlen: emitted as a singleton (outs) — too short to have room for a mate.
- readlen <= residual < 2×readlen: read 1 is a full readlen read; the remainder becomes its mate if that remainder is >= minlen (an uneven-length pair), otherwise read 1 becomes a singleton and the sub-minlen tail goes to discards.
Output headers carry tab-delimited provenance back to the source molecule:
- Pairs:
shredpb_<n> <1|2>:N:0 originalHeader pairStartInMolecule— the<1|2>:token lets BBTools auto-detect interleaving, and<n>is a global pair index shared by both reads of a pair. - Singletons:
shredpb_s<n> originalHeader startInMolecule - Discards:
shredpb_d<n> originalHeader startInMolecule
Support
For questions and support:
- Email: bbushnell@lbl.gov
- Documentation: bbmap.org
- Related tool: Shred — the general-purpose single-end shredder this tool is derived from.