We plan genomics projects and build the software that runs them.
Project planning, bioinformatics pipelines, LIMS/QMS for NGS and arrays, and tertiary analysis. Each one is built for your project and deployed on your servers or in your cloud.
We scope a genomics project before any samples are run: what to sequence or genotype, on which platform, at what depth, and what it takes to analyse and validate it.
We work out
Assay and platform choice: NGS panel, exome, genome or array
Coverage, sample size and batch design
Compute and storage sizing
Validation plan and accreditation requirements
You receive
A written project plan with timeline and budget lines
Analysis specification the pipeline will be built to
We do not ship one generic pipeline. Each build starts from your project scope, includes only the modules that scope needs, and is validated before we hand it over.
Tailored to project scope
Short-read alignment
Long-read alignment
Array genotype calling
Duplicate and UMI handling
Germline SNV and indel calling
Somatic SNV and indel calling
Copy number
Structural variants and fusions
Coverage and call-rate QC
Sample identity and sex check
Validated before handover
Reference samplesSamples with known truth sets run through the finished pipeline
ConcordanceCalls compared with the truth set, variant class by variant class
Acceptance limitsResults checked against the limits agreed in your project plan
Signed reportValidation report issued, then the pipeline is handed over
SensitivityPass
PrecisionPass
ReproducibilityPass
Illustrative layout. The white marker is the acceptance limit set for your project; a pipeline is released only when every metric clears it.
Snakemake or Bash · pinned conda environments · local or cloud
A laboratory information and quality management system for NGS and array laboratories. Every sample, bench step and QC decision is recorded with a signature and timestamp.
Pipelines that turn variant calls into an interpreted report for a specific clinical application. They run locally, so patient data stays on your site.
Scope
Annotation against versioned public sources
Rule-based interpretation from a curated knowledge base
Select a stage to see what goes in, what runs and what comes out.
Goes in
FASTQ or BAM/CRAM; array intensity files
What runs
Alignment, variant calling or genotyping, then coverage, mapping and variant QC for every run and sample.
Comes out
BAM, VCF and QC tables checked against your acceptance limits
Start a project
Build your project brief
Pick what applies. The brief on the right updates as you go; copy it and send it to us.
project_brief.txt
Paste it into an email to mg@mgbioscience.com. We reply with times for a 30-minute technical call.
Built with this approach
Two tertiary analysis platforms
OncoSight
Clinical decision support for oncology. It brings clinical and molecular results together and presents the evidence relevant to the case. The treating oncologist makes the decision.
Input
Clinical data and molecular test results
Output
Case report for the oncologist, summary for the patient
NutriSight
Nutrition and lifestyle reporting from genotype data, interpreted against a curated nutrigenomics knowledge base.
Input
Genotypes from microarray, NGS or targeted qPCR
Output
Personal report in HTML and JSON, with a practitioner viewer