Bioinformatics analysis service

Genomics analysis services: SNV, indel, CNV, and SV calling from short- and long-read DNA

Genomics analysis turns DNA sequencing reads into annotated SNV and indel calls using GATK Best Practices [1]. CNV, SV, and long-read modules can be added at kickoff. Pepkio uses version-pinned pipelines, documented scripts, publication figures, and a Methods draft for academic, biotech, and pharma teams. Custom inputs, outputs, and non-standard workflows are agreed at kickoff.

Key facts

Key facts about genomics & variant analysis analysis
FactValue
Data types supportedIllumina short-read paired-end FASTQ or BAM; MGI DNBSEQ-T7 / G400 / G99 when scoped at kickoff; Element Biosciences AVITI and Ultima UG100 when scoped at kickoff; PacBio and Oxford Nanopore long-read FASTQ for the long-read spoke; pre-aligned BAM or gVCF for variant-calling-only projects
Reference builds or standards usedHuman GRCh38 primary with GATK resource bundle 4.6 [2]; GRCh38 alternate loci per functional-equivalence standards [3]; legacy GRCh37/hg19; mouse GRCm39; custom references scoped at kickoff
Primary tools (with versions)BWA-MEM2 2.2.1; GATK 4.6.0.0; Picard 3.2.0; samtools/bcftools 1.21; mosdepth 0.3.3; Ensembl VEP 112; Manta 1.6.0; DELLY 1.2.6; minimap2 2.28; Clair3 1.0.9; fastp 0.23.4; FastQC 0.12.1; MultiQC 1.25; DeepVariant 1.8.0 optional
Typical turnaround range2-4 weeks (single-sample germline); 4-8 weeks (multi-sample cohort with joint genotyping and VQSR), confirmed at kickoff
Deliverable formats.bam, .g.vcf.gz, filtered .vcf.gz, annotation .tsv, CNV/SV call tables; PDF/SVG figures; HTML QC report; commented bash/R/Python scripts; Methods draft
Regulatory/reproducibility standards followedGATK Best Practices [1]; functional-equivalence alignment standards [3]; MD5 checksum validation; version-pinned software with documented parameters; clinical validation guidance where applicable [4, 5]

What is genomics?

Genomics studies an organism's complete DNA sequence and the computational analysis used to detect variation against a reference genome. A single-nucleotide variant (SNV) is a one-base change; an indel is a short insertion or deletion; a copy-number variant (CNV) is a gain or loss of genomic segments; and a structural variant (SV) is a larger rearrangement such as an inversion or translocation. Typical questions include which DNA sequence changes relate to disease risk, drug response, or evolutionary divergence. The NIH All of Us Research Program released 245,350+ whole-genome sequences to researchers by 2023, with planned releases exceeding 400,000 WGS [10].

What genomics analysis can answer

Genomics analysis identifies inherited and somatic DNA changes linked to disease mechanism, treatment response, and population genetic architecture. Applications include rare undiagnosed disorders, cancer driver catalogs, and cohort-scale association studies [4, 11]. The examples below pair research questions with published results.

Which pathogenic variants explain an undiagnosed rare disease when coding panels and exome sequencing are inconclusive?

Clinical WGS validation guidelines emphasize detecting noncoding variants and structural changes that WES may miss [4].

Which somatic driver mutations are actionable across a large cancer cohort?

Kinnersley et al. [11] analyzed 10,478 cancer WGS profiles across 35 tumor types, identifying 330 candidate driver genes; approximately 55% of patients harbor at least one clinically relevant mutation.

Does germline TP53 loss of heterozygosity precede clinical cancer diagnosis in predisposed individuals?

Light et al. [12] reported near-ubiquitous early TP53 LOH with gain of the mutant allele years before tumor diagnosis in Li-Fraumeni syndrome WGS data.

Which CNVs and SVs underlie developmental disorders when SNV/indel panels are negative?

Gabrielaite et al. [13] benchmarked 11 CNV callers on matched WES and WGS data and suggested combining GATK gCNV, Lumpy, DELLY, and cn.MOPS, since no single caller in that comparison captured all CNV classes reliably.

How can population cohorts share harmonized variant calls for association studies?

Regier et al. [3] defined functional-equivalence standards, including BWA-MEM alignment, GRCh38 with alternate loci, and CRAM compression, that reduce cross-center call discordance.

Services included in this category

Genomics & Variant Analysis services offered by Pepkio
ServiceDescriptionPrimary tools
Whole-genome sequencingGenome-wide germline SNV and indel discovery from raw FASTQs with callable-region QCBWA-MEM2, GATK HaplotypeCaller, Ensembl VEP
Whole-exome sequencingTargeted coding-region variant calling from capture-enriched libraries at lower sequencing cost than WGSBWA-MEM2, GATK, VEP; capture-aware depth QC
Variant callingSNV/indel calling, joint genotyping, and filter tuning when alignment is already completeGATK HaplotypeCaller, DeepVariant, bcftools
CNV and structural variationGenome-wide copy-number and structural variant detection from WGS or WES alignmentsGATK gCNV, Manta, DELLY; CNVkit when scoped
Long-read DNA sequencingPhased variants and structural events from PacBio or Oxford Nanopore readsminimap2, Clair3; long-read SV callers when scoped

What Pepkio delivers

Genomics projects include filtered variant files, coverage QC tables, annotated figures, version-pinned scripts, and a Methods draft. Extra outputs can be added at kickoff. Standard deliverables include:

Variant files

  • Indexed .bam
  • .g.vcf.gz
  • Filtered .vcf.gz
  • variant_annotation_master.tsv with consequence and population/clinical fields when configured

QC tables and figures

  • Coverage summaries, mapping and depth metrics, FastQC/MultiQC reports
  • Coverage distributions, Ti/Tv and consequence plots
  • VQSR tranche curves when applicable

CNV/SV outputs when scoped

  • Segment tables and SV VCF with breakpoint evidence

Code and documentation

  • Commented bash/R/Python scripts with lock files
  • README and Methods draft
  • Post-delivery reviewer clarification for analyses we performed

How the analysis works — step by step

  1. 1. Validate inputs and sample metadata

    Verify FASTQ or BAM integrity (MD5 checksums), read layout, read groups, and experimental design; record sample_manifest.csv and flag sub-threshold yield.

    Tools and outputs

    Tools used: md5sum; custom validation scripts

  2. 2. QC and trim raw reads

    FastQC 0.12.1 and fastp 0.23.4 assess and trim reads; flag low Q30 or high-adapter libraries before alignment.

  3. 3. Align reads to the reference genome

    BWA-MEM2 2.2.1 aligns short reads to GRCh38; minimap2 2.28 on the long-read spoke; samtools 1.21 indexes coordinate-sorted BAMs.

  4. 4. Mark duplicates and recalibrate base qualities

    Picard 3.2.0 MarkDuplicates and GATK 4.6.0.0 BaseRecalibrator / ApplyBQSR using the GATK resource bundle.

  5. 5. Assess coverage and callability

    mosdepth 0.3.3 and Picard CollectWgsMetrics or CollectHsMetrics report mean coverage and breadth at agreed thresholds [4].

  6. 6. Call SNVs and indels

    GATK 4.6.0.0 HaplotypeCaller in gVCF mode; GenomicsDBImport and GenotypeGVCFs for cohorts; DeepVariant 1.8.0 optional [8].

  7. 7. Call CNVs and structural variants when scoped

    GATK gCNV, Manta 1.6.0, DELLY 1.2.6, and CNVkit when scoped; Gabrielaite et al. [13] suggested combining multiple CNV methods.

  8. 8. Annotate, filter, and package deliverables

    Ensembl VEP 112 annotation; GATK VQSR or hard filters [2]; bcftools 1.21; MultiQC 1.25; final scripts, README, Methods draft, and figures.

Tools and standards we use

Core stack for short-read and long-read genomics projects. Versions are pinned at kickoff and cited in the Methods draft.

Genomics & Variant Analysis tools and standards
ToolVersionRolePrimary citation
BWA-MEM22.2.1Short-read alignment to GRCh38[6] — https://doi.org/10.1109/IPDPS.2019.00041
GATK4.6.0.0Duplicate marking, BQSR, variant calling, VQSR[1] — https://doi.org/10.1002/0471250953.bi1110s43
Picard3.2.0MarkDuplicates, coverage metrics[1] — https://doi.org/10.1002/0471250953.bi1110s43
samtools / bcftools1.21BAM indexing, VCF manipulationLi et al., 2009 — https://doi.org/10.1093/bioinformatics/btp352
mosdepth0.3.3Fast coverage depth and breadthPedersen & Quinlan, 2018 — https://doi.org/10.1101/517435
Ensembl VEP112Variant consequence annotation[7] — https://doi.org/10.1186/s13059-016-0974-4
DeepVariant1.8.0Neural-network SNV/indel caller (optional)[8] — https://doi.org/10.1038/nbt.4235
Manta1.6.0SV discovery from paired readsChen et al., 2016 — https://doi.org/10.1093/bioinformatics/btv710
DELLY1.2.6SV and indel callingRausch et al., 2012 — https://doi.org/10.1093/bioinformatics/bts378
minimap22.28Long-read alignment[3] — https://doi.org/10.1093/bioinformatics/bty191
Clair31.0.9Long-read SNV/indel callingZheng et al., 2022 — https://doi.org/10.1038/s41467-022-32121-6
MultiQC1.25Aggregated QC reporting[9] — https://doi.org/10.1093/bioinformatics/btw354

Common challenges — and how we handle them

Teams outsourcing genomics analysis often run into pipeline reproducibility, caller disagreement, CNV/SV detection variability, storage cost, and undocumented filters [5, 13, 14].

Pipeline choice can affect reproducibility more than sequencing platform.

Pan et al. [14] tested 56 aligner-caller combinations and found pipelines had a larger impact on inherited-variant reproducibility than platform or library prep. Pepkio pins versions and documents parameters.

Inherited calls can disagree across pipelines.

Pan et al. [14] found SNV reproducibility varies across pipelines; complementary filters or callers can reduce disagreement. Pepkio applies documented GATK or DeepVariant filters and discusses multi-caller strategies when needed.

CNV/SV detection varies by assay and tool.

Gabrielaite et al. [13] found WGS outperforms WES and suggested combining GATK gCNV, Lumpy, DELLY, and cn.MOPS. Pepkio scopes multi-tool workflows at kickoff.

Storage and transfer costs grow at cohort scale.

Regier et al. [3] showed CRAM reduces a 30× WGS BAM from ~54 Gb to ~17 Gb (>3-fold). Pepkio delivers CRAM or compressed VCF where agreed.

Filters are hard to reproduce if they are not documented.

Koboldt [5] emphasizes rigorous filtering and benchmarking. Pepkio exports VCF FILTER labels and drafts Methods with exact versions.

Common questions

What data do I need to provide for a genomics analysis project?

Provide paired-end FASTQ or coordinate-sorted BAMs, sample metadata, and a target reference build. Clinical WGS guidelines recommend agreeing on mean coverage and callability metrics at kickoff [4], often around 30x depth with over 95% callability. MD5 checksums are requested at transfer. Non-standard inputs are scoped at kickoff.

How long does genomics analysis take at Pepkio?

Single-sample germline projects typically take 2-4 weeks; multi-sample cohorts with joint genotyping and VQSR typically take 4-8 weeks. CNV/SV or long-read add-ons can extend timelines. Milestones are confirmed at kickoff.

What do the deliverables look like?

Deliverables include filtered VCFs, annotation tables, coverage QC summaries, PDF/SVG figures, MultiQC HTML, commented scripts with lock files, and a Methods draft. Bespoke formats such as plink exports, IGV sessions, and custom gene lists are agreed at kickoff.

Can you handle my specific sequencing platform or instrument?

Illumina NovaSeq X, 6000, NextSeq 2000, and HiSeq use BWA-MEM2 + GATK. MGI DNBSEQ-T7, G400, and G99 FASTQs are processed when scoped at kickoff with adapter/QC validation in the report. Element AVITI and Ultima UG100 are processed when scoped at kickoff. PacBio Revio and Oxford Nanopore route to the long-read spoke (minimap2, Clair3).

What if my data quality is poor or depth is below recommended thresholds?

Sub-threshold libraries are flagged before calling. Low Q30, high adapter content, or depth below agreed thresholds can reduce callable fraction [4]. Pepkio discusses re-sequencing versus stricter filters before proceeding.

Do you provide the analysis code, and do I own it?

Yes. You retain full ownership. Pepkio delivers commented bash, R, and Python scripts with lock files via private Git or agreed transfer, with notebooks available on request.

Can I be involved during the analysis?

Yes. Checkpoint reviews occur after alignment QC, after coverage assessment, and before delivery. A PhD-level bioinformatician is your primary contact.

What happens if a journal reviewer requests changes after delivery?

Clarification of methods, QC thresholds, and minor figure or table revisions for analyses we performed are included. Substantial new reviewer-requested analyses are scoped separately.

Should I use GRCh38 or hg19 for my genomics analysis?

Pepkio defaults to GRCh38 with GATK bundle 4.6 [2] and alternate loci [3]. GRCh37/hg19 is supported on request for legacy cohorts.

Can Pepkio handle custom or non-standard genomics analyses?

Yes. Client BAMs, custom references, plink exports, somatic extensions, and workflows outside the five spokes are scoped at kickoff with documented deliverables.

Related services

  • Transcriptomics — Link germline or somatic variants to gene expression for eQTL mapping, allele-specific expression, and multi-omics interpretation.
  • Custom consulting — Pre-sequencing depth, cohort size, reference-build, and capture-kit planning before library prep.
  • Whole-genome sequencing — Detailed FASTQ-to-VCF workflow when genome-wide coverage is your primary need.
  • CNV and structural variation — Dedicated CNV/SV calling when SNV/indel analysis is already complete or insufficient for your biological question.
References
  1. Van der Auwera GA, Carneiro MO, Hartl C, et al. From FastQ data to high-confidence variant calls: the Genome Analysis Toolkit best practices pipeline. Current Protocols in Bioinformatics. 2013;43(1):11.10.1–11.10.33. https://doi.org/10.1002/0471250953.bi1110s43 (PMID: 25431634)
  2. Broad Institute. GATK 4.6.0.0 release notes and VQSR documentation. 2024. https://github.com/broadinstitute/gatk/releases/tag/4.6.0.0; https://gatk.broadinstitute.org/hc/en-us/articles/360035531612-Variant-Quality-Score-Recalibration-VQSR
  3. Regier AA, Farjoun Y, Larson DE, et al. Functional equivalence of genome sequencing analysis pipelines enables harmonized variant calling across human genetics projects. Nature Communications. 2018;9:4038. https://doi.org/10.1038/s41467-018-06159-4 (PMID: 30279509)
  4. Rehm HL, Bale SJ, Bayrak-Toydemir P, et al. Best practices for the analytical validation of clinical whole-genome sequencing intended for the diagnosis of germline disease. npj Genomic Medicine. 2021;6(1):47. https://doi.org/10.1038/s41525-020-00154-9 (PMID: 33110627)
  5. Koboldt DC. Best practices for variant calling in clinical sequencing. Genome Medicine. 2020;12:91. https://doi.org/10.1186/s13073-020-00791-w (PMID: 33106175)
  6. Vasimuddin M, Misra S, Li H, Aluru S. Efficient architecture-aware acceleration of BWA-MEM for multicore systems. IEEE IPDPS. 2019. https://doi.org/10.1109/IPDPS.2019.00041
  7. McLaren W, Gil L, Hunt SE, et al. The Ensembl Variant Effect Predictor. Genome Biology. 2016;17(1):122. https://doi.org/10.1186/s13059-016-0974-4 (PMID: 27268795)
  8. Poplin R, Chang PC, Alexander D, et al. A universal SNP and small-indel variant caller using deep neural networks. Nature Biotechnology. 2018;36(10):983–987. https://doi.org/10.1038/nbt.4235 (PMID: 30247488)
  9. Ewels P, Magnusson M, Lundin S, et al. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics. 2016;32(19):3047–3048. https://doi.org/10.1093/bioinformatics/btw354 (PMID: 27312411)
  10. NIH All of Us Research Program. Program update — Denny J. 2025. https://dpcpsi.nih.gov/sites/default/files/2025-01/Day-1-215PM-All-of-Us-Program-Update-Denny-v3-508.pdf
  11. Kinnersley B, Sud A, Everall A, et al. Analysis of 10,478 cancer genomes identifies candidate driver genes and opportunities for precision oncology. Nature Genetics. 2024;56(9):1868–1877. https://doi.org/10.1038/s41588-024-01785-9 (PMID: 38890488)
  12. Light N, Layeghifard M, Attery A, et al. Germline TP53 mutations undergo copy number gain years prior to tumor diagnosis. Nature Communications. 2023;14:77. https://doi.org/10.1038/s41467-022-35727-y (PMID: 36604421)
  13. Gabrielaite M, Torp MH, Rasmussen MS, et al. A comparison of tools for copy-number variation detection in germline whole exome and whole genome sequencing data. Cancers. 2021;13(24):6283. https://doi.org/10.3390/cancers13246283 (PMID: 34944901)
  14. Pan B, Kusko R, Xiao W, et al. Assessing reproducibility of inherited variants detected with short-read whole genome sequencing. Genome Biology. 2022;23:2. https://doi.org/10.1186/s13059-021-02569-8 (PMID: 34980216)

Individual services

Deep-dive pages for specific genomics & variant analysis methods and workflows.

Let's Talk About Your Science

Tell us:

  • • Your biological question
  • • Data type and size
  • • Timeline constraints

We'll tell you:

  • • What's feasible
  • • How long it will take
  • • Exactly what it will cost
Discuss Your Project

Contact us to start with a free consultation. Need everyday bench calculators? Try our free lab tools.