PacBio Sequencing Data Analysis — Turn Revio HiFi Reads into Publication-Ready Results
PacBio HiFi sequencing, generated on the current Revio system, produces reads with a median length of 15-20 kb and consensus accuracy above 99.9%, giving researchers unbiased, near-complete coverage of the genome, transcriptome, or epigenome in a single dataset. Raw HiFi reads only become useful once they are aligned, called, assembled, and annotated—and PacBio's toolchain has evolved substantially since the Sequel-generation era. CD Genomics' PacBio sequencing data analysis service runs the current, community-standard toolchain—pbmm2, DeepVariant, pbsv, HiPhase, TRGT, HiFiCNV, hifiasm, and the Iso-Seq analysis suite—to turn Revio HiFi data into variant calls, assemblies, isoforms, and methylation maps ready for interpretation or publication.
Whether your HiFi data comes from our own sequencing service or was generated elsewhere, our dedicated bioinformatics team builds a workflow around your specific question—human, plant, animal, or microbial—rather than running a one-size-fits-all pipeline.
Why researchers choose our PacBio data analysis service
- Current, actively maintained toolchain—no legacy CLR-era software
- High-precision SNV/indel calling with DeepVariant's PacBio model
- Structural variant, tandem repeat, and copy number detection from a single dataset
- Haplotype-resolved de novo assembly with hifiasm or verkko
- Full-length isoform and 5mC methylation analysis from the same HiFi reads
At a glance:
- Introduction
- What Is PacBio Sequencing Data Analysis?
- Data Sources We Analyze
- Key Advantages of Our Analysis Service
- Applications
- Analysis Workflow - How It Works
- Analysis Modules and Tools
- Choosing the Right Analysis Module
- Data Input Requirements
- Why Choose CD Genomics
- Case Study: A Reproducible Pipeline for PacBio WGS and Repeat Expansion Analysis
- FAQs
- Demo
Introduction
Long-read HiFi sequencing on the PacBio Revio system generates far more than raw reads—within a single dataset lie SNVs, indels, structural variants, tandem repeat expansions, copy number changes, haplotype phase, full-length transcript isoforms, and CpG methylation status. Extracting each of these signals reliably requires the right tool at each step, applied with the right parameters for your organism and question. Tool choices that were standard in the Sequel II era—built around continuous long reads (CLR) rather than HiFi consensus reads—no longer reflect current best practice. CD Genomics keeps its PacBio analysis pipelines current with the tools PacBio and the long-read community actively maintain and benchmark today.
What Is PacBio Sequencing Data Analysis?
PacBio sequencing data analysis is the set of computational steps that convert raw HiFi reads into interpretable biological results. Starting from unaligned BAM files produced by the Revio system, reads are demultiplexed, aligned to a reference (or assembled de novo when no reference exists), and passed through variant calling, phasing, structural variant, tandem repeat, and—where relevant—transcript or methylation analysis modules.
Because HiFi reads already carry near-perfect per-base accuracy from the sequencer, this analysis focuses on structural and biological interpretation rather than error correction, which was a major burden with older continuous long-read (CLR) data.
Our team runs this analysis using PacBio SMRT sequencing data as the primary input, and can integrate results alongside Oxford Nanopore-derived datasets for projects that combine both platforms.
Data Sources We Analyze
Our pipelines are built around HiFi data from the current PacBio instrument generation, with support for legacy datasets where needed.
PacBio Revio HiFi Data (Primary)
The Revio system, running SPRQ-Nx chemistry, is our primary data source. Revio HiFi reads carry a median consensus accuracy of Q30 or better, with on-instrument 5mC (and 5hmC) methylation calling already embedded in the aligned BAM. We align this data with pbmm2, a minimap2 wrapper purpose-built for HiFi reads.
Legacy Sequel/Sequel II Datasets
For customers with existing Sequel or Sequel II HiFi (CCS) data, we can still run the modern variant-calling and assembly toolchain, though we recommend Revio for new sequencing given its higher throughput and lower per-genome cost.
Data-analysis-only engagements are welcome—send us your existing HiFi BAM or FASTQ files. If you also need new sequencing, we offer PacBio library construction and Revio sequencing through our PacBio Pre-Made Library Sequencing service.
Key Advantages of Our Analysis Service
Scientific Advantages
- Current, benchmarked toolchain
pbmm2, DeepVariant, pbsv, HiPhase, TRGT, and HiFiCNV form our default WGS variant workflow, matching the pipeline PacBio itself recommends for Revio data.
- Haplotype-resolved output
HiPhase jointly phases small variants, structural variants, and tandem repeats using HiFi read-backed phasing, giving allele-specific results without a separate assay.
- Assembly without a reference
hifiasm and verkko deliver telomere-to-telomere-quality, haplotype-resolved de novo assemblies directly from HiFi reads, with optional Hi-C or trio-binning integration.
- Multi-omic from one dataset
Full-length isoform identification (Iso-Seq) and CpG methylation profiling can both be run on the same HiFi reads used for variant calling, maximizing the value of each SMRT Cell.
Business & Project Advantages
- Flexible entry point
Bring your own HiFi data, or combine sequencing and analysis as a single project—both are supported.
- Organism-agnostic pipelines
The same core toolchain scales from human clinical-research samples to plant, animal, and microbial genomes, with organism-specific reference and parameter tuning.
- Reproducible, containerized workflows
Our pipelines run in containerized, version-controlled environments, so results are reproducible and auditable for publication.
- Clear, interpretable reports
Deliverables include annotated VCFs, assembly statistics, isoform tables, and methylation tracks, alongside a plain-language summary report.
Applications
Whole-Genome Variant Detection
- High-precision SNV and indel calling with DeepVariant's PacBio-trained model.
- Structural variant detection (insertions, deletions, inversions, duplications, translocations) with pbsv.
- Tandem repeat genotyping and visualization with TRGT, including repeat-expansion loci relevant to neurological disease research.
- Copy number variant detection with HiFiCNV and joint phasing with HiPhase.
De Novo Genome Assembly
- Haplotype-resolved assembly with hifiasm, alone or combined with Hi-C or parental short-read data.
- Gapless, telomere-to-telomere-style assembly with verkko for reference-grade genome projects.
Full-Length Transcriptome Analysis (Iso-Seq)
- Full-length isoform identification and quantification from HiFi cDNA or Kinnex data.
- Novel isoform and fusion transcript discovery using current classification tools.
Epigenetic and Complex-Population Analysis
- On-instrument 5mC/5hmC methylation calling from native HiFi reads, no bisulfite conversion required.
- Minor variant and base-modification detection in bacterial, fungal, and viral populations.
Analysis Workflow - How It Works
1. Pre-Processing
Unaligned HiFi BAM files from the Revio system are demultiplexed if needed, then aligned to a reference genome with pbmm2, or routed directly to assembly if no reference exists. Reads are sorted and indexed for downstream tools.
2. Variant Calling and Phasing
SNVs and small indels are called with DeepVariant; structural variants with pbsv; tandem repeats with TRGT; and copy number variants with HiFiCNV. HiPhase then jointly phases all variant types using HiFi read-backed information.
3. Assembly or Transcript/Methylation Analysis
For de novo projects, hifiasm or verkko builds haplotype-resolved contigs directly from HiFi reads. For transcriptomic projects, full-length reads are classified into isoforms. For epigenetic projects, on-instrument 5mC/5hmC calls are extracted and mapped to genomic context.
4. Annotation and Reporting
- Variants are annotated against relevant databases and filtered by quality and population frequency where applicable
- Assembly statistics (N50, BUSCO completeness, phasing quality) are calculated
- Isoform tables and methylation tracks are generated for visualization in IGV or similar tools
- A summary report and a MultiQC-style run report are compiled for every project
Workflow of PacBio HiFi data analysis, from raw Revio reads through alignment or assembly, variant calling and phasing, to annotated, publication-ready output.
Analysis Modules and Tools
| Analysis Module | Current Toolchain | Detects | Typical Applications |
| Whole-genome variant calling | pbmm2, DeepVariant, pbsv, TRGT, HiFiCNV, HiPhase | SNVs, indels, SVs, tandem repeats, CNVs, phased haplotypes | Human, plant, animal, and microbial genome research |
| De novo assembly | hifiasm, verkko | Haplotype-resolved, near-gapless genome assemblies | Reference genome construction, pan-genome projects |
| Full-length RNA analysis (Iso-Seq) | SMRT Link Iso-Seq pipeline, isoform classification and quantification tools | Full-length isoforms, novel transcripts, fusion events | Transcriptome annotation, alternative splicing studies |
| Epigenetics and base modification | On-instrument 5mC/5hmC calling, SMRT Link kinetics tools | CpG methylation, DNA/RNA base modification, motif analysis | Allele-specific methylation, epigenetic regulation studies |
| Complex population analysis | pbmm2, DeepVariant, pbsv, minor-variant calling workflows | De novo assembly, minor variant and modification detection | Bacterial, fungal, and viral population studies |
Choosing the Right Analysis Module
Most projects combine more than one module. The table below summarizes what each core module adds, to help you scope your project.
| Module | Reference Genome Needed? | Best For | Not Ideal For |
| WGS variant pipeline | Yes | SNV/indel/SV/CNV detection, phasing, repeat expansion genotyping | Organisms without a usable reference genome |
| De novo assembly (hifiasm/verkko) | No | New reference genomes, non-model organisms, pan-genome studies | Quick variant screening against a known reference |
| Iso-Seq transcript analysis | Recommended, not strictly required | Full-length isoform discovery, alternative splicing, fusion detection | DNA-level structural variant detection |
| Methylation/kinetics analysis | Yes | Allele-specific methylation, epigenetic regulation | Samples without matched genomic context |
How to interpret this comparison
- Start with the WGS variant pipeline when a reference genome is available and your priority is comprehensive variant detection with phasing.
- Choose de novo assembly when no adequate reference exists, or when you need a new haplotype-resolved reference for your species.
- Add Iso-Seq or methylation modules on top of either pipeline when your project also requires transcript-level or epigenetic information from the same samples.
Data Input Requirements
| Category | Requirement | Notes |
| File format | Unaligned HiFi BAM (preferred) or FASTQ | Native PacBio BAM preserves kinetics tags needed for methylation calling |
| Instrument generation | PacBio Revio (preferred); Sequel/Sequel II HiFi (CCS) also supported | CLR-only data requires additional preprocessing and is not recommended for new projects |
| Minimum coverage - WGS variant calling | ≥ 15-20× for germline SNVs/indels; ≥ 30× recommended for full SV/CNV/phasing confidence | Coverage needs scale with genome size and heterozygosity |
| Minimum coverage - de novo assembly | ≥ 20× HiFi coverage, higher for highly heterozygous or polyploid genomes | Hi-C or trio data further improves haplotype resolution |
| Reference genome | Provided by customer, or selected jointly from public databases | Required for alignment-based modules; not required for de novo assembly |
| Data transfer | Secure cloud transfer or physical hard drive | We can also receive data directly if sequencing was performed by CD Genomics |
Why Choose CD Genomics
Current, Community-Aligned Toolchain
We run the same core tools—pbmm2, DeepVariant, pbsv, HiPhase, TRGT, HiFiCNV, hifiasm—that define current best practice for Revio HiFi data analysis.
Flexible Engagement
Analyze data you already have, or combine it with our Revio sequencing and library construction services for an end-to-end project.
Multi-Omic Expertise
From variant calling to de novo assembly, Iso-Seq, and methylation analysis, our bioinformaticians work across the full range of HiFi applications.
Organism-Agnostic Experience
Human, plant, animal, and microbial genome projects are all supported, with reference and parameter choices tailored to your organism. Structural variant work often pairs well with our dedicated human genome structural variation detection service.
Transparent, Reproducible Reporting
Every project is delivered with containerized, version-tracked workflows and a clear summary report suitable for methods sections and internal review.
Case Study: A Reproducible Pipeline for PacBio WGS and Repeat Expansion Analysis
Jain, T., Clelland, C. nf-core/pacvar: a pipeline for analyzing long-read PacBio whole genome and repeat expansion sequencing data. Bioinformatics 41(4), btaf116 (2025).
1. Background
PacBio long-read sequencing enables both whole-genome variant detection and targeted characterization of complex repeat-expansion loci linked to neurodegenerative disease, through PacBio's PureTarget panel. Existing toolkits such as SMRT Link and PacBio's HiFi-human-WGS-WDL pipeline each carried infrastructure limitations—SMRT Link could not run natively on macOS, and the WDL pipeline lacked compatibility with certain HPC schedulers and cloud platforms. There was a clear need for a single, portable pipeline covering both use cases.
2. Methods
The authors built nf-core/pacvar, a Nextflow pipeline with three main components:
- Pre-processing: demultiplexing with Lima, alignment with pbmm2, sorting/indexing with SAMtools
- Variant calling and phasing: SNV calling with DeepVariant or HaplotypeCaller, structural variant calling with pbsv, and joint phasing with HiPhase
- Repeat expansion characterization: genotyping and visualization of 20 neurological-disease-associated repeat loci with TRGT
The pipeline was built to run identically across HPC clusters with different schedulers, cloud platforms, and local systems, using Docker, Singularity, Apptainer, or conda containers.
3. Results
The nf-core/pacvar pipeline structure, comprising pre-processing, variant calling and phasing, and repeat expansion characterization, with the specific tool used at each step.
Key Findings
- Dual variant-calling support:
- The pipeline allows users to choose between DeepVariant and HaplotypeCaller for SNV calling, and supports concordance analysis between the two callers on the same sample.
- Joint phasing across variant types:
- HiPhase phases small variants, structural variants, and tandem repeats together, which the authors highlight as important for interpreting methylation and CRISPR-editing outcomes in an allele-specific manner.
- Infrastructure-independent deployment:
- Because the pipeline is built on Nextflow and nf-core's standardized template, it runs across HPC schedulers, cloud environments, and local systems without the platform restrictions of earlier toolkits.
- Standardized quality reporting:
- Every run generates a MultiQC report summarizing execution runtime, resource usage, and successful completion, supporting reproducibility across projects.
- Flexible, modular design:
- Users can skip SV calling, SNV calling, or phasing steps independently, tailoring the pipeline to WGS-only or repeat-expansion-only projects.
4. Conclusions
This study demonstrates that a modular, containerized pipeline built on current PacBio HiFi tools can deliver both comprehensive WGS variant detection and specialized repeat-expansion characterization in a single, portable workflow. Importantly:
- pbmm2, DeepVariant, pbsv, and HiPhase together form a coherent, current-generation variant-calling toolchain for HiFi data
- Joint phasing of small variants, structural variants, and repeats adds interpretive value beyond calling each variant type in isolation
- TRGT-based repeat genotyping supports characterization of clinically relevant repeat-expansion loci directly from WGS or targeted HiFi data
- Containerized, standardized pipelines improve reproducibility and portability across computing environments
FAQs
Q: Do I need to sequence with CD Genomics to use this analysis service?
No. We accept HiFi BAM or FASTQ data generated on your own Revio, Sequel, or Sequel II system. If you also need new sequencing, we can combine library construction and Revio sequencing with the analysis in a single project.
Q: What is the difference between the WGS variant pipeline and de novo assembly?
The WGS variant pipeline aligns your reads to an existing reference genome to call SNVs, indels, structural variants, tandem repeats, and CNVs. De novo assembly builds a new genome sequence directly from your reads, without a reference—useful for non-model organisms or when a haplotype-resolved reference genome is the goal itself.
Q: Can you call DNA methylation from my existing HiFi data?
Yes, provided the original BAM files retain PacBio's kinetics tags. Revio data includes on-instrument 5mC/5hmC calling by default; older Sequel data may require reprocessing depending on how it was generated.
Q: Which structural variant and small variant callers do you use?
Our default pipeline uses DeepVariant for SNVs and small indels and pbsv for structural variants, aligned with pbmm2 and jointly phased with HiPhase. This mirrors the current best-practice toolchain recommended for PacBio Revio HiFi data.
Q: Can this service analyze repeat expansion loci relevant to neurological disease research?
Yes. Using TRGT-based genotyping, we can characterize repeat number, motif composition, and mosaicism at known repeat-expansion loci, whether from whole-genome data or targeted repeat-expansion sequencing.
Q: How much HiFi coverage do I need?
For germline SNV/indel calling against a reference, 15-20× coverage is typically sufficient, with 30× recommended for confident structural variant, CNV, and phasing results. De novo assembly projects generally need 20× or more, with additional Hi-C or trio data improving haplotype resolution.
Q: Can you analyze microbial or complex population samples?
Yes. We support de novo assembly and minor-variant/base-modification detection in bacterial, fungal, and viral populations, using the same core alignment and variant-calling toolchain adapted for population-scale data.
Q: What do I receive at the end of the project?
Deliverables typically include annotated variant call files, assembly statistics or isoform/methylation tables as relevant, a MultiQC-style run report, and a plain-language summary suitable for internal review or a methods section.
Demo
1. Phased Small Variant, Structural Variant, and Tandem Repeat Overview
2. Haplotype-Resolved Assembly Contiguity Summary
3. Repeat Expansion Genotyping Waterfall Plot
References
- Jain, T., Clelland, C. nf-core/pacvar: a pipeline for analyzing long-read PacBio whole genome and repeat expansion sequencing data. Bioinformatics. 41(4), btaf116 (2025).
- Wenger, A.M., Peluso, P., Rowell, W.J. et al. Accurate circular consensus long-read sequencing improves variant detection and assembly of a human genome. Nat Biotechnol. 37, 1155-1162 (2019).
- Cheng, H., Concepcion, G.T., Feng, X. et al. Haplotype-resolved de novo assembly using phased assembly graphs with hifiasm. Nat Methods. 18, 170-175 (2021).
- Holt, J.M., Saunders, C.T., Rowell, W.J. et al. HiPhase: jointly phasing small, structural, and tandem repeat variants from HiFi sequencing. Bioinformatics. 40, btae042 (2024).
For Research Use Only. Not for use in diagnostic procedures.