Scroll to navigation

QUIVER(1) QUIVER(1)

NAME

quiver - genomic consensus caller designed for Pacific Biosciences data

DESCRIPTION

Quiver is an algorithm for calling highly accurate consensus from multiple PacBio reads, using a pair-HMM exploiting both the basecalls and QV metrics to infer the true underlying DNA sequence.

Data file requirements

To make the most accurate consensus calls possible, Quiver makes use of a battery of quality value metrics calculated by the basecaller. If you are using a SMRTportal installation verision 1.4 or later, then SMRTportal will load all the information Quiver needs, so you can skip the rest of this section.
In SMRTportal versions 1.3.3 and prior, by default only a subset of these quality values are included in the .cmp.h5 files produced by SMRTanalysis. To get a .cmp.h5 with all the QVs loaded, you will need to use the RS_Mapping_QVs protocol to create a cmp.h5 file for Quiver.
If you are using an older version than SMRTportal/SMRTanalysis 1.3.3, please upgrade.

EXAMPLES

Running Quiver

For example,
$ quiver -j8 aligned_reads.bam              \
>        -r path/to/lambda.fasta            \
>        -o variants.gff -o consensus.fasta


will use 8 CPUs to run Quiver on aligned_reads.bam, outputting the consensus sequence and variants.
Note that if you are using a cmp.h5 file and have not used the RS_Mapping_QVs protocol to generate that file---or if the source bas.h5 file was generated by pre-1.3.1 instrument software---the cmp.h5 will not contain the full battery of QV metrics required for optimal Quiver accuracy. The command will still work, but it will give a warning that its accuracy will be suboptimal.

Run Quiver on Multiple Input Files

Multiple alignment files in a FOFN (File of File Names) can be quivered against a single reference ( GenomicConsensus >= 1.1.0).
An example input FOFN:
$ cat aligned_reads.fofn
/path/to/reads1.bam
/path/to/reads2.bam


can be used instead of a reads file:
$ quiver -j8 aligned_reads.fofn             \
>        -r path/to/lambda.fasta            \
>        -o variants.gff -o consensus.fasta


Quiver can also be used with DataSet XML files. See pbcore for details on generating new DataSet XML files for your alignment files.

Highly-accurate assembly consensus

Quiver enables consensus accuracies on genome assemblies at accuracies approaching or even exceeding Q60 (one error per million bases). If you use the HGAP assembly protocol in SMRTportal 2.0 or later, Quiver runs automatically as the final "assembly polishing" step.

SEE ALSO

quiver-faq(7) plurality(1) variantCaller(1) pbgff(5)
February 2016