Bioinformatics & genomics

Turning genomicdata intomeaningful insight.

We build bioinformatics solutions that help transform complex genomic data into usable insights, workflows and applications.

  • Genomic data analysis
  • Variant interpretation
  • Multi-omics integration
  • Data pipelines
  • Genomics visualisation
Talk to our teamSee how we build

Genomic sequence

ATGCGATCTAAG
GCTACGATGTCC
TTACGGTACTGA

Expression analysis

Variant analysis

  • SNP
  • Indel
  • CNV
  • Structural variant

Multi-omics

  • Genomics
  • Transcriptomics
  • Proteomics
  • Metabolomics
  • ISO/IEC27001
  • ISO/IEC42001
  • DPA andBAA ready
  • AWS, Azure,Google Cloud
  • All cloudregions

OrbitNexa builds and validates genomic analysis pipelines for clinical labs, diagnostics centres, biotech and research institutes: whole-exome and genome secondary analysis with ACMG/AMP classification, metagenomics, isolate genomics and RNA-Seq, as containerised, version-pinned Nextflow workflows with a full provenance trail, run in the client's own cloud account in India or the UK, aligned to NABL, ISO 15189 and IVDR validation expectations.

The catalogue

Every pipeline, every tool version, every reference, and whether it is built or in development.

Eight analyses built across three pipeline families; wider clinical, research and epigenomics scope in development, and said to be so.

  • Germline WES / WGS

    From raw reads to an ACMG/AMP-classified, tiered clinical report. Built.

    BWA-MEM2 or Parabricks → GATK HaplotypeCaller or DeepVariant → VEP / SnpEff → ClinVar, gnomAD, OMIM → ACMG/AMP classification → tiered report

    Built · GIAB truth sets · MANE Select transcripts

    GATKParabricksVEPClinVar

  • Microbiome & metagenomics

    16S profiling, shotgun functional profiling, metagenome-assembled genomes. Built.

    QIIME2 / DADA2 with SILVA · Kraken2, MetaPhlAn, HUMAnN · MEGAHIT or metaSPAdes · CheckM · diversity and differential abundance

    Built · versioned reference databases per run

    QIIME2Kraken2MetaPhlAnNextflow

  • Isolate genomics & surveillance

    Bacterial WGS with AMR and MLST, fungal assembly, amplicon species ID. Built.

    SPAdes, Bakta, AMRFinderPlus / CARD, MLST, Snippy · AMR surveillance and wastewater metagenomics for public-health labs

    Built · One Health surveillance

    SPAdesBaktaCARDDocker

  • Transcriptomics

    Bulk RNA-Seq expression analysis across short-read and Nanopore direct-RNA data. Built.

    STAR or Salmon · DESeq2 / edgeR · nf-core/rnaseq conventions · single-cell (Cell Ranger, Seurat / Scanpy) in development

    Built · single-cell in development

    STARSalmonDESeq2Nextflow

  • Somatic, PGx, long-read & more

    The wider clinical scope, in development and labelled as such.

    Somatic panels with Mutect2, CNV and fusion callers, COSMIC and GENIE · pharmacogenomics with star alleles and CPIC · PacBio HiFi and ONT with minimap2 and Clair3 · NIPT · methylation

    In development · ACMG/AMP v4-ready classification with a transition mode

    Mutect2Clair3CPICPython

  • LIMS, EHR & ABDM integration

    Order-to-report automation once the interfaces are mapped.

    HL7 v2 orders and results · FHIR Genomics observations and diagnostic reports, vcf2fhir · HGVS-clean structured reports · ABDM health-record linkage · LIMS connectors

    ABDM-FHIR · NABL 112 workflow alignment

    FHIRHL7 v2FastAPIPostgreSQL

Bioinformatics & Genomics

Reads to report.

  • Validated
  • Pinned
  • Replayable
  • Provenance recordEvery version, every run.

Our process

From assay scope to a validated pipeline in your account.

A validation dossier as much as a service: truth sets, per-variant metrics, re-validation triggers and a statement of limitations, written the way an assessor expects.

  1. 01

    Scope

    Discovery & Strategy

    1-2 weeks

    We map your ecosystem, constraints and KPIs before any engineering starts, so every technical decision has a reason on record.

  2. 02

    Design

    Architecture & Design

    2-3 weeks

    Architects draw the system and designers prototype the interface, both reviewed and signed off before a line of code is written.

  3. 03

    Build

    Agile Development

    4-12 weeks

    Iterative sprints on modern frameworks, with a senior engineer reviewing every pull request before it merges.

  4. 04

    Validate

    Quality Assurance

    2-4 weeks

    Automated unit, integration and acceptance tests, plus performance and security checks against real-world scenarios.

  5. 05

    Operate

    Launch & Evolution

    Ongoing

    A zero-downtime deployment, then continuous monitoring and iterative enhancement as your standing technical partner.

How we validate.

Truth sets, per-variant metrics with confidence intervals, re-validation triggers, and a statement of limitations in every report.

  1. Truth setsGIAB reference samples per assay; in-silico data only as a supplement.
  2. MetricsSensitivity and positive predictive value with 95% confidence intervals for SNV, indel, CNV and SV; limit of detection for somatic.
  3. ReproducibilityInter-run and inter-operator concordance. Identical inputs give identical outputs.
  4. Re-validationTriggered by any tool, reference or database version change; supplemental validation documented before release. ClinVar appraised quarterly, every annotation resource yearly.
  5. TraceabilityReference genome, transcript set (MANE), database versions and container digests recorded per run and printed in the report, with a statement of what the assay does not detect.

Who is on the engagement.

Overlap with UK and Indian working hours every day, a shared channel, a weekly call and a validation review with your quality lead every two weeks.

  • Engagement leadA senior bioinformatician, not an account manager
  • Two pipeline engineersNextflow, containers, cloud batch
  • Clinical genomics specialistInterpretation logic, ACMG/AMP, report design
  • Platform engineerYour account, provenance store, cost ledger
  • Your quality lead and cliniciansOwn the validation plan and the sign-off
The full process, phase by phase

Let’s build together

Have an idea?
Let’s make it real.

Tell us about your goals. We’ll help you find the right way forward.

  • Share your idea

    Tell us what you're looking to build.

  • Explore possibilities

    We'll understand your goals and suggest the right approach.

  • Plan the next steps

    Together we define the roadmap.

  • Build what's next

    Turn ideas into real impact.

Let’s discuss
your project

Whether it’s a new product, a platform upgrade or a complex challenge, we’re here to help.

Get in touch

Engagement models

Start by validating one analysis on your samples.

Shapes described by what happens, how long they run and what you hold at the end. Never a per-sample price.

  • Pipeline Validation

    One analysis built and validated on your samples against reference data, with QC thresholds agreed first.

    TeamLead, pipeline engineer, your quality lead
    Timeline4 – 8 weeks
    Key deliverablesA validated pipeline and a validation report you can show an accreditation assessor

  • Platform Build

    Pipelines in production in your cloud account, with the sign-off workflow, provenance store and interoperability wired in.

    TeamLead, two pipeline engineers, clinical specialist, platform engineer
    Timeline10 – 20 weeks
    Key deliverablesA genomics platform your lab runs and your auditor can replay

  • Bioinformatics Retainer

    Pipelines kept current with references and tools, new assays added, and a review every month.

    TeamA named bioinformatician, on the platform
    TimelineOngoing
    Key deliverablesA monthly record of what ran and how, and the V&V package kept current

Send us a FASTQ

One sample, no commitment

Tell us the assay, the platform, the sample count and the region. You get a QC report (FastQC / MultiQC, mosdepth coverage), a variant or profile summary, and the turnaround and compute usage recorded for that run.

Send a sample

Where your samples' data lives.

Your account, an India or UK region, on-prem where required, and a ledger you can read.

  • Your account, your region

    AWS Mumbai or Hyderabad, Azure Central India, Google Cloud Mumbai, or UK regions. On-prem or Slurm HPC where required, with the same provenance record.

  • DPDP posture stated

    Genetic data handled as sensitive; cross-border transfer on the negative-list basis; federated analysis and trusted-research-environment patterns for multi-site studies.

  • You see the ledger

    Compute and storage attributed per sample and per pipeline, tagged by project, with spot and GPU strategy stated and tiered archiving from FASTQ to CRAM.

  • The V&V package is yours

    Validation plan and report, IEC 62304 lifecycle records, ISO 14971 risk file, 21 CFR Part 11 evidence, change control and CAPA procedure.

Accreditation, mapped

NABL 112, ISO 15189, CAP, IVDR, FDA, DPDP and ABDM, against what we deliver.

We align to these; the lab holds the accreditation. What we deliver is the documentation and the pipeline behaviour each one expects.

What each framework asks for, and what we deliver

  • NABL 112 and ISO 15189:2022 — the pipeline documented as part of the test procedure, every component approved by the lab director, end-to-end validation with human samples, four unique identifiers per data file
  • CAP checklist and the AMP/CAP guideline — variant-type-specific metrics, supplemental validation on any significant change, reference-genome version tracked per software version, proficiency testing readiness
  • IVDR and FDA — per-component validation from base calling to reporting, IEC 62304 lifecycle at Class B or C, an ISO 14971 risk file, the bioinformatics validation section of the technical documentation
  • DPDP and ABDM — genetic data as sensitive data, India region pinning, consent and purpose limitation, ABDM-FHIR interoperability
  • GxP and 21 CFR Part 11 — audit trails, electronic signatures, chain of custody per sample

QC before compute is spent

  • Q30 yield, mean and target coverage with thresholds by assay (15x / 20x / 30x)
  • Duplication rate, contamination estimate with rejection at the stated threshold, insert size, on-target rate
  • Tumour purity threshold for somatic assays
  • A MultiQC panel per batch, kept with the provenance record

AI in interpretation.

Explainable, evidence-grounded, human-signed.

  1. 01A tool per variant class — AlphaMissense for missense, SpliceAI for splicing, Evo 2 embeddings on ClinVar — never one model for everything.
  2. 02Every suggestion carries the retrieved evidence behind it, so a clinician reviews the reasoning, not a score.
  3. 03A clinician signs before any classification is final. We build the gate; we do not stand in it.
  4. 04Validated under the same V&V package as the pipeline, and under our ISO/IEC 42001 AI management system.

FAQ

Questions?
We're here to help.

Truth sets, ACMG/AMP v4, database updates, HPC, GPU, LIMS. Straight answers with the thresholds named. Still have a question? We're just a message away.

Get in touch
  1. 01Which sequencing platforms and assays do you support?
    Short-read data from Illumina and MGI and long-read data from Oxford Nanopore. Built today: whole-exome secondary analysis with ACMG/AMP classification, 16S and shotgun metagenomics with metagenome-assembled genomes, bacterial and fungal isolate genomics, and bulk RNA-Seq. Wider clinical, research and epigenomics scope is in development, and we say which is which.
  2. 02Is the pipeline reproducible enough for an accreditation audit?
    That is what it is built for. Every run is containerised and version-pinned, with the tool versions, reference genome and database versions recorded per sample, so identical inputs give identical outputs and any run can be replayed. The validation report is written against the metrics your quality lead names, in the form a NABL assessor expects.
  3. 03Can patient data stay in India?
    Yes. Pipelines run in your own AWS, Azure or Google Cloud account in an India region, with personal data handled under the DPDP Act and the audit trail kept in-region alongside it.
  4. 04Do you provide the clinical interpretation, or only the pipeline?
    The pipeline, the annotation and a draft classification with its evidence assembled per variant. A clinician signs before a report exists; we build the sign-off gate, we do not stand in it. The report is tiered, NABL-aligned and exportable as FHIR.
  5. 05How is this related to InferaGen.ai?
    Same team, same pipelines. InferaGen.ai is the clinical genomics platform we are building ourselves, in active development, with eight analyses built across three pipeline families. Client work on this line uses the same engineering and the same provenance discipline, in the client's own account.
  6. 06Which truth sets and thresholds do you validate against?
    GIAB reference samples, with sensitivity and positive predictive value per variant type and 95% confidence intervals; germline coverage thresholds per assay are written into the validation plan.
  7. 07Are you ready for ACMG/AMP v4?
    The classifier supports points-based scoring and ClinGen specifications, with a transition mode that reports both versions and a migration plan for historical calls.
  8. 08How do you handle a ClinVar or reference update?
    Databases are pinned per run. Updates are appraised on a stated cadence (ClinVar quarterly, all annotation resources yearly), re-validated, and released as a new pipeline version with the change recorded.
  9. 09Can the pipeline run on our own HPC?
    Yes. The same Nextflow workflows run on Slurm with Singularity or Apptainer containers, with the same provenance record.
  10. 10Do you support somatic, pharmacogenomics, long-read or single-cell?
    They are listed in the catalogue with their status: built lines are validated, in-development lines are said to be so.
  11. 11How is cost visible without a quote?
    Every run is tagged; compute and storage are attributed per sample and per pipeline, and the ledger is yours.
  12. 12Can you integrate with our LIMS and hospital systems?
    HL7 v2 and FHIR Genomics, HGVS-clean structured reports, ABDM linkage, and order-to-report automation once the interfaces are mapped.
  13. 13What about GPU acceleration?
    NVIDIA Parabricks where the science permits and the cost profile favours it; DRAGEN where the lab already runs it; CPU GATK as the reference path.
  14. 14How does the AI in variant interpretation stay safe?
    Evidence-grounded suggestions, a tool per variant class, clinician sign-off before anything is final, validated under the same package as the pipeline and under ISO/IEC 42001.

From the work on this line

Real problems. Real outcomes.

View all insights

Have a genomics or multi-omics workflow to build?

Book a 30-minute call with the team building the pipelines. We'll review your assays and tell you what a validated pipeline would take.

Talk to a bioinformaticianEmail hello@orbitnexa.com
+91 912-195-7728Hyderabad, IndiaEvery brief gets a senior review. Reply within 1 business hour, 9 AM-7 PM IST.