Sequel IIe (SQIIe) FAQ

Is ccs the same for on-instrument Sequel IIe and off-instrument SMRT Link?

Yes. The source code is identical for a given release, e.g. ICS v10.0 and SMRT Link v10.0. The only difference is how we generate the ccs binaries for SQIIe and SMRT Link; each binary is compiled for different environments, with varying degrees of hardware specializations and optimizations. ccs binaries for SQIIe are slightly faster than for SMRT Link, but all binaries from a given release yield identical results.

Will it be like this in the future?

There are no plans to bifurcate. In fact, we dedicate a lot of energy into having all x86 binaries compile from the same source code and produce identical results.

What is the difference between bioconda and official PacBio releases?

Bioconda releases are only available as developer versions. They may contain features and bug fixes that have not been integrated in our release software yet.

Which ccs options are used for on-instrument processing?

SQIIe default: --all

SQIIe with kinetic information: --all --all-kinetics --subread-fallback

But the `reads.bam` header contains significantly more options!

Yes, those are mostly used to explicitly name auxilliary files. Everything explained:

/opt/pacbio/pa-ccs/current/bin/ccs \ # binary
    /data/pa/m64012_201202_224049.consensusreadset.xml \ # XML output name
    --all --all-kinetics --subread-fallback\ # all mode with kinetics
    --streamed \ # do not use a BAM file as input, but a BAM input stream
    --log-level INFO \
    --suppress-reports \ # do not generate report or metric files per default
    --log-file /data/pa/m64012_201202_224049.ccs.log \ # log file name
    --bam /data/pa/m64012_201202_224049.reads.bam \ # explicit BAM file name
    --report-json /data/pa/m64012_201202_224049.ccs_reports.json \ # the ccs_reports file as JSON
    --report-file /data/pa/m64012_201202_224049.ccs_reports.txt \ # the ccs_reports file as human readable
    --metrics-json /data/pa/m64012_201202_224049.zmw_metrics.json.gz \ # metrics file name
    --hifi-summary-json /data/pa/m64012_201202_224049.hifi_summary.json \ # summary JSON file for hifi statistics
    --stderr-json-log # instrument-specific additional JSON logging

What files (formats) are transferred from the instruments?

For HiFi runs, the main file is a reads.bam file, prefixed by the movie name, and it is accompanied by two core auxilliary files:

*.ccs_reports.log
*.zmw_metrics.json.gz

What is in the `reads.bam`?

The on-instrument ccs version and also SMRT Link ≥v10 run in the --all mode by default. In this mode, ccs outputs one representative sequence per productive ZMW, irrespective of quality and passes. More information in the --all mode FAQ and in the reads.bam FAQ.

Can you go back to `subreads.bam` from `reads.bam`?

Not when operating the instrument in CCS mode. See next question.

Can I get subreads and CCS reads transferred for the same SMRT Cell?

You can select either CLR or CCS reads coming off the instrument. If you want CLR and CCS reads, you can select HiFi generation on SMRT Link, which transfers CLR off instrument and runs CCS on your SMRT Link instance:

What is the expected run time for ccs on the instrument and will it delay the next SMRT cell?

The design of SQIIe allows to have ccs running in parallel with the acquisition of the subsequent SMRT cell, and will not interfere or delay subsequent acquisitions. The hardware and software is configured to accommodate future increases in HiFi yield.

How does ccs on- and off-instrument performance compare?

Performance of ccs on the instrument is roughly equivalent to running on a dedicated system with 384 logical cores. On-instrument processing includes parallel post-primary computation and result transfers up to 90% faster than for CLR runs.

Can you specify custom parameters on the instrument, like by-strand CCS?

There is no possibility to specify custom parameters. We are open to feedback what might be of general interest in future instrument software updates.

Can you include kinetics information or is it included by design?

You may select to retain kinetic information in Run-Design:

Be aware that the output BAM files will more than 5x larger due to additional kinetics.

Can I generate a `reads.bam` from a `subreads.bam` file which was generated with older chemistry / instrument software (even from the Sequel System)?

Generating a reads.bam has been possible since ccs v5 and supported chemistries can be found here.

What is the file size of a HiFi dataset transferred from the instruments for one SMRT Cell on the SQIIe System?

We provide upper bounds and average file sizes how the BAM file size scales with HiFi yield in the BAM output FAQ. From this page, a typical 30 GBases HiFi yield run will result in a ~33 GByte reads.bam file with default parameters and ~174 GB with HiFi kinetics included.

Can demultiplexing of barcoded datasets be done on the SQIIe instrument?

No, this currently not supported. Demultiplexing continues to be supported through SMRT Link or with lima on the command line.

Can the `reads.bam` still be used in a CLR application? E.g. to make full use of all the data, not just the HiFi reads?

In SMRT Link and in general, only the HiFi reads are used from the reads.bam. It should not be necessary to include other reads. Some pipelines do allow including CCS reads <Q20 (e.g. Iso-Seq). You are free to use the all reads from the reads.bam files with your own tools.

Can I extract the different subtypes of reads from the new `reads.bam`?

Yes, the SMRT Analysis pipeline “Export Reads” in SMRT Link v10.0 or newer can export HiFi reads to BAM/FASTA/FASTQ format; when adjusting minimum CCS predicted accuracy, you can include CCS reads <Q20. On the command line, tools can be used to filter the BAM file for the read quality rq tag, please see the reads.bam FAQ.