Whole genomes in under six hours.Every result checked against truth.
GenoAccel runs FPGA-accelerated secondary analysis of next-generation sequencing data on AWS F2 instances — returning aligned reads, called variants and quality metrics, with the concordance record published rather than claimed.
Definition
What GenoAccel is
The speedup means different things to different buyers, which is why one engine serves two markets. A research core facility experiences it as cost per genome and throughput. A neonatal intensive care unit experiences it as a diagnosis arriving inside the window where it can still change treatment.
- GenoAccel
- GenoAccel is a managed genomics analysis platform that runs FPGA-accelerated secondary analysis of next-generation sequencing data on AWS F2 instances. Research and clinical laboratories upload raw sequencing reads in FASTQ format, and GenoAccel returns aligned reads, called variants, quality-control metrics and interpretable reports — targeting a 30× whole human genome in under six hours against a CPU baseline of roughly sixteen hours.
02 · Measured, not marketed
Four numbers, each with its conditions attached.
A statistic without provenance is noise. Sample, genome build, reference set and pipeline version travel in the row with the number they qualify.
| Metric | Value | Sample | Build | Reference set | Pipeline | State |
|---|---|---|---|---|---|---|
| 30× WGS wall clock | — | NA12878 | GRCh38 | — | — | Pending |
| SNV F1 score | — | NA12878 | GRCh38 | — | — | Pending |
| Indel F1 score | — | NA12878 | GRCh38 | — | — | Pending |
| Cost per genome | — | — | — | — | — | Pending |
Why every value is an em-dash. No figure is published here until it has actually been measured on real hardware. An honest gap beats a plausible invention, and this audience checks.
03 · How it works
The page is the pipeline.
Each segment is a real process with a digest-pinned tool version, so the run that produced a result can be reproduced exactly.
- 01
Upload
Presigned multipart transfer straight to object storage. Resumable, checksum-verified.
- 02
Validate
FASTQ format, paired-end pairing, read length and quality encoding, before compute is spent.
- 03
Align
BWA-MEM alignment against GRCh38.
FPGA - 04
Call variants
Variant calling to VCF.
FPGA - 05
Quality control
Coverage, duplication and threshold evaluation. Below-threshold samples are flagged.
- 06
Deliver
Variants, metrics and artifacts — queryable, paginated, exportable.
04 · Accuracy
The concordance record, published either way.
Every pipeline version is validated against Genome in a Bottle reference materials, and the result is published — including the runs that did not improve on the previous version.
| Pipeline | Truth set | SNV F1 | Indel F1 | Run date | State |
|---|---|---|---|---|---|
| — | NA12878 / HG001 | — | — | — | Not yet run |
| — | NA24385 / HG002 | — | — | — | Not yet run |
05 · Trust
What is in place, and what is not.
Dates rather than adjectives. A control appears as in place only when the artifact evidencing it exists.
| Control | State | Evidence |
|---|---|---|
| AWS Business Associate Addendum | In place | AWS Artifact |
| Tenant isolation at the data layer | In place | Cross-tenant tests |
| PHI redaction in logs | In place | Automated test |
| Encryption in transit and at rest | In place | TLS 1.3, KMS |
| SOC 2 Type II | Roadmap | Window not started |
| CLIA / CAP certification | Roadmap | Gates clinical use |
06 · Questions
Frequently asked questions
What is GenoAccel?
GenoAccel is a managed genomics analysis platform. Research and clinical laboratories upload raw next-generation sequencing data, and GenoAccel runs FPGA-accelerated secondary analysis on AWS F2 instances, returning aligned reads, called variants, quality-control metrics and reports.
How does FPGA acceleration make genomic analysis faster?
Field-programmable gate arrays are well suited to the specific computational shape of read alignment and variant calling. AWS cites up to 30× NGS compute acceleration for FPGAs versus current CPU instances. Counting input/output and the pipeline stages that do not accelerate, GenoAccel targets a conservative 3–5× end-to-end improvement on a full whole-genome pipeline, against a CPU baseline of roughly 16 hours for a 30× whole human genome.
Is GenoAccel HIPAA compliant, and can it be used for clinical diagnostics?
GenoAccel operates under an AWS Business Associate Addendum and restricts protected health information to HIPAA-eligible services. Clinical diagnostic use additionally requires CLIA certification, which in the United States is held by the laboratory reporting the result. GenoAccel supports a certified laboratory's compliance by providing versioned pipelines, immutable results and audit trails. SOC 2 Type II is on the roadmap and is not claimed until the report exists.
What sequencing data formats and assay types does GenoAccel accept?
GenoAccel accepts FASTQ input, validated for format, paired-end pairing, read length and quality encoding before compute is spent. Supported assay types are whole-genome sequencing, whole-exome sequencing, targeted gene panels and RNA sequencing, each with its own pipeline profile and quality thresholds.
How does GenoAccel verify that its variant calls are accurate?
Every pipeline version is run against Genome in a Bottle reference materials, and the resulting concordance record is published in full, including runs that did not improve on the previous version. The validation suite runs monthly and on every pipeline change. Figures are published only after they have been measured.
How large a file can be uploaded to GenoAccel?
Uploads use presigned multipart transfer directly to object storage, with part size scaled to the file so a transfer stays within the protocol's 10,000-part limit. A 100 GB whole-genome FASTQ set is a supported case. Interrupted uploads resume by re-sending only the missing parts, and the assembled object is verified against the SHA-256 checksum and byte count declared before the transfer began.
Who is GenoAccel for?
Academic core facilities and research laboratories, biotechnology companies and contract research organisations, and clinical laboratories. A research laboratory experiences the speedup primarily as reduced cost per genome, while a clinical setting such as a neonatal intensive care unit experiences it as a diagnosis arriving inside the window where it can still change treatment.