Analysis-Only Bioinformatics · Biomedical · Agrigenomics

Bring your data — or generate it with us.

You do not need to be a sequencing client. Provide raw data from any source — your own lab, a collaborator, a public repository — or generate it in-house with Nucleome, and our bioinformatics team delivers publication-grade analysis with reproducible, versioned pipelines.

6
omics layers, one analysis team
3
named pipelines: OmniIntegrate · Neoantigen · siRNA
Versioned
reproducible workflows in every project
NDA
confidential data handling

From first principles to expert-level analysis

New here? Start with Foundations. Planning a study or reviewing results? Jump to Practitioner or Expert. Each level builds on the one before it, so you can read this page as a short course or dip in where you need to.

Start here if you are new to this service — six ideas that explain what it is and why it matters.

LEVEL 1

What is bioinformatics analysis?

The computational interpretation of biological data — turning raw sequencing, mass-spectrometry, or array files into biological answers: which variants, which genes, which pathways, which candidates.

LEVEL 1

What does "analysis-only" mean?

You supply the raw data (from your lab, a collaborator, or a public repository) or generate it with Nucleome — we perform the computational analysis.

LEVEL 1

What is OmniIntegrate?

Nucleome's multi-omics integration pipeline: it jointly analyses two or more omics layers (e.g. RNA-seq + proteomics + metabolomics) to find shared biology, integrated pathways, and predictive models.

LEVEL 1

What is neoantigen discovery?

Identifying tumour-specific peptide candidates that could be recognised by the immune system — from tumour sequencing data through HLA-binding prediction — for vaccine and immunotherapy research.

LEVEL 1

What is siRNA discovery?

Designing small interfering RNAs against a target gene with minimal off-target risk — for knockdown experiments and therapeutic-development research.

LEVEL 1

What will I get?

Processed results (variant calls, count matrices, protein tables, candidate lists), publication-ready figures, a written report with methods, and reproducible pipeline documentation.

For scientists planning a project or reading a report: how each step works and how it can mislead.

LEVEL 2

Data intake and QC

Raw data is verified (checksums, integrity), quality-controlled (FastQC-class metrics, MS QC), and any issues are reported before analysis proceeds.

LEVEL 2

Pipeline execution

Documented, versioned workflows — BWA/GATK-class variant calling, STAR/Salmon-class transcriptomics, MaxQuant/FragPipe-class proteomics — with parameters reported in the deliverable.

LEVEL 2

Statistical analysis

Differential analysis with multiple-testing correction, enrichment analysis, and supervised models only with proper validation — we do not publish over-fitted PLS-DA or unvalidated classifiers.

LEVEL 2

Multi-omics integration (OmniIntegrate)

Latent-factor models (MOFA+/DIABLO-class), joint pathway mapping, and correlation networks — with permutation testing and cross-validation for any predictive model.

LEVEL 2

Neoantigen pipeline steps

Somatic variant calling → expression confirmation → HLA typing → binding prediction (NetMHCpan-class) → candidate ranking with evidence tiers — predictions clearly labelled as predictions.

LEVEL 2

siRNA design steps

Target-transcript analysis → guide design with seed-region and isoform specificity → SNP/variant avoidance → off-target scoring → ranked candidate list for empirical validation.

For study design: the decisions that determine whether the data answers your question.

LEVEL 3

Power and design before analysis

Sample size, batch structure, and confounders determine what any analysis can detect — we advise at design stage, before data exists.

LEVEL 3

Reproducibility standards

Pinned versions, containerised workflows, and full provenance — analyses can be regenerated and defended in review.

LEVEL 3

Neoantigen validation reality

Binding predictions are predictions: only a fraction of predicted binders are immunogenic. We tier candidates by evidence strength and flag what requires empirical validation.

LEVEL 3

siRNA off-target reality

Seed-mediated off-targeting is unavoidable in principle; design minimises it, and empirical validation (multiple independent guides, rescue experiments) remains essential.

LEVEL 3

Machine learning with integrity

Predictive models are trained with proper cross-validation, external validation where possible, and honest reporting of limitations — no leakage, no overfitting.

LEVEL 3

Public-data reanalysis

TCGA, GEO, ENCODE, and similar resources are reanalysed with full provenance — useful for validation cohorts and hypothesis generation.

OmniIntegrate
Nucleome's multi-omics integration pipeline (latent-factor modelling, joint pathways, ML).
Neoantigen
a tumour-specific peptide candidate for immune recognition, predicted from sequencing data.
siRNA
small interfering RNA; guides sequence-specific mRNA knockdown.
MOFA+/DIABLO-class
latent-factor and multi-omics integration frameworks.
NetMHCpan-class
peptide–MHC binding prediction tools.
FDR
false discovery rate; the multiple-testing standard for omics differential analysis.

Bioinformatics Analysis Services at Nucleome Informatics

Nucleome Informatics provides data-analysis-only bioinformatics services through two divisions: Biomedical and Agrigenomics: raw data must be supplied by the client or generated by Nucleome. Every analysis runs on documented, versioned pipelines with methods, parameters, and software versions reported in the deliverable.

  • Reproducible analysis with documented software versions, parameters, and reference data
  • Publication-ready figures and tables
  • A written analysis report with methods, results, and interpretation notes
  • Raw and processed output files (VCF, count matrices, feature tables, etc.)
  • Consultation on study design before analysis begins — the cheapest place to fix a study is at design

Analysis portfolio

  • Genomics: WGS/WES variant calling (SNV/indel, CNV, SV), annotation (VEP/ANNOVAR-class), ACMG-style classification support, trio and cohort analysis, GWAS and fine-mapping.
  • Transcriptomics: bulk RNA-seq (alignment/quantification, differential expression, pathway analysis), single-cell RNA-seq (QC, clustering, trajectory, cell–cell communication), spatial transcriptomics.
  • Epigenomics: ChIP-seq, ATAC-seq, bisulfite/methylation analysis.
  • Metagenomics: 16S/ITS amplicon and shotgun analysis, MAG assembly, taxonomic and functional profiling.
  • Proteomics & metabolomics: database-search reanalysis, differential abundance, PTM analysis, pathway mapping.
  • Multi-omics integration — OmniIntegrate pipeline: joint analysis of two or more omics layers (genomics + transcriptomics + proteomics + metabolomics) — MOFA+/DIABLO-class factor analysis, joint pathway mapping, correlation networks, and machine-learning models with honest validation.
  • Neoantigen discovery pipeline: somatic mutation calling from tumour WGS/WES → RNA-seq expression confirmation → HLA typing → peptide–MHC binding prediction (NetMHCpan-class tools) → prioritised neoantigen candidate list with evidence tiers.
  • siRNA discovery pipeline: target-gene siRNA design with off-target minimisation (seed-region and transcript-isoform specificity analysis, SNP and variant avoidance), delivery-strategy notes, and ranked candidate guides for empirical validation.
  • Custom pipeline development: bespoke workflows, Nextflow/Snakemake-class reproducibility, and deployment for your team.

Data handling

  • Client raw data is analysed as provided; we verify data integrity (checksums, QC) before analysis and report any issues found.
  • Data confidentiality is maintained under NDA; data is not reused or shared.
  • Public-data reanalysis (TCGA, GEO, ENCODE, and similar) is supported with full provenance.

Biomedical and Agrigenomics data analysis

Our bioinformatics services are organised into two divisions, each with its own specialists and pipelines, sharing the same versioned, reproducible workflows.

Human & clinical research

Biomedical Data Analysis Services

For hospitals, pharma, biotech and academic groups working on human disease, cancer and therapeutics.

  • Multi-omics integration (OmniIntegrate): joint analysis of genome, transcriptome, proteome and metabolome data with latent-factor models and validated machine learning
  • Neoantigen discovery: tumour WGS/WES + RNA-seq → HLA typing → pMHC binding prediction → evidence-tiered candidates
  • siRNA discovery: off-target-minimised guide design with isoform specificity and SNP avoidance
  • Biomarker discovery: differential analysis, feature selection and cross-validated classifiers for diagnostic, prognostic and predictive markers
  • Single-cell & spatial transcriptomics: QC, clustering, cell-type annotation, trajectories, cell–cell communication and spatial domains
  • Clinical & cancer genomics: germline and somatic WGS/WES variant calling, CNV/SV, ACMG-style classification support
  • Transcriptomics & epigenomics: bulk RNA-seq, fusions, splicing, ChIP/ATAC-seq and methylation
  • Human microbiome: 16S and shotgun metagenomics for gut, oral and skin studies
Discuss a biomedical project →
Plant · Animal · Microbial

Agrigenomics Data Analysis Services

For crop and livestock breeders, agricultural universities, ICAR institutes and agri-biotech companies.

  • T2T genome assembly: telomere-to-telomere, gap-free chromosomes from long reads (HiFi/ONT ultra-long) with Hi-C scaffolding and haplotype phasing
  • De novo genome assembly: draft-to-chromosome-level assemblies with polishing, BUSCO/QV completeness checks, and gene and repeat annotation
  • Microbial genome assembly: complete bacterial and fungal genomes, plasmid resolution, AMR/virulence genes and pangenomes
  • RNA-seq data analysis: differential expression in crops, livestock and non-model species; de novo transcriptome assembly; co-expression networks
  • Population genomics & breeding: SNP discovery, GBS/ddRAD genotyping, diversity and structure, GWAS and QTL mapping, genomic selection support
  • Comparative & pan-genomics: synteny, gene-family evolution, phylogenomics and pan-genome graphs
  • Plant, animal & microbial multi-omics: integrating genome, transcriptome, metabolome and phenotype data with OmniIntegrate
  • Environmental & agri-metagenomics: soil, rhizosphere, rumen and aquaculture microbiomes; MAG assembly and functional profiling
Discuss an agrigenomics project →

Every service, standard and advanced

Analysis-only: raw data must be supplied by the client or generated by Nucleome. The table maps each service, by division, to what its standard and advanced tiers cover.

ServiceStandard analysisAdvanced analysis
Biomedical data analysis
WGS / WESAlignment, SNV/indel calling, variant annotation, QC reportsCNV/SV calling, trio de novo analysis, homozygosity mapping, ACMG-style classification support
Bulk RNA-seqAlignment/quantification, differential expression, PCAPathway enrichment, alternative splicing, fusion detection, cell-type deconvolution
Single-cell RNA-seqQC, clustering, cell-type annotation, marker genesTrajectory and pseudotime, cell–cell communication, multi-sample integration
Spatial transcriptomicsSpot/cell QC, clustering, spatial expression plotsCell-type deconvolution, spatial ligand–receptor analysis, tissue-domain detection
Epigenomics (ChIP / ATAC / methylation)Peak calling, differential binding or methylationMotif enrichment, regulatory-network linkage, multi-layer integration
Metagenomics (16S / shotgun)Taxonomic profiling, diversity analysisMAG assembly, functional profiling, strain-level analysis
ProteomicsProtein identification and quantitation, differential abundancePTM site analysis, interactomics scoring, pathway enrichment
MetabolomicsFeature-table processing, differential abundancePathway-interaction networks, isotope-tracer and flux analysis
GWAS / genotypingQC, association testing, Manhattan and QQ plotsFine-mapping, polygenic risk scores, heritability, ancestry-aware models
Multi-omics (OmniIntegrate)Joint PCA/UMAP, correlation networksMOFA+/DIABLO-class factor models, validated machine-learning classifiers
Neoantigen discoverySomatic variant calling, HLA typing, binding predictionsExpression-aware evidence tiering, immunogenicity features, vaccine-candidate ranking
siRNA discoveryGuide design with off-target scoringIsoform-specific design, SNP/variant avoidance, ranked candidate panels
Agrigenomics data analysis (plant · animal · microbial)
T2T / de novo genome assemblyLong-read assembly, polishing, BUSCO and QV completeness, basic annotationHi-C scaffolding to chromosome or T2T level, haplotype-phased assemblies, full gene and repeat annotation
Microbial genome assemblyComplete genome assembly, annotation, AMR and virulence genesPlasmid resolution, pangenome and comparative genomics
Plant & animal RNA-seqReference-based quantification and differential expressionDe novo transcriptome assembly, co-expression networks (WGCNA-class), alternative splicing
Population genomics & breedingSNP calling, genetic diversity and population structureGWAS and QTL mapping, genomic selection support, selection-signature scans
Comparative & pan-genomicsSynteny and orthology analysisGene-family evolution, phylogenomics, pan-genome graphs
Agri-metagenomicsTaxonomic profiling of soil, rhizosphere or rumen samplesMAG assembly, functional and pathway profiling
Both divisions
Custom pipeline developmentBespoke workflow development for your dataContainerised deployment, documentation, and team handover

"Standard" covers the core analysis that answers the primary question for each data type; "Advanced" covers deeper methods that need more project context and are scoped at the design consultation. Tier selection is confirmed per project before analysis begins — no analysis starts without an agreed scope.

How the service works, step by step

Multi-omics integration frameworks (MOFA+, DIABLO), peptide–MHC binding prediction (NetMHCpan), and siRNA design principles are established in the literature — see references below.

Figure 1 Analysis-only service model. schematic
Client-provided raw data Nucleome-generated data Any source: your lab, collaborator, public repository Integrity-checked at intake Analysed as provided, under NDA In-house or partner lab Same versioned pipelines Full provenance end to end
Figure 1 | Analysis-only service model. (Illustrative schematic, not sample data.)
Figure 2 OmniIntegrate workflow. schematic
Design consultation Data intake & QC Versioned pipeline execution Statistics & integration Figures & interpretation Reproducible report
Figure 2 | OmniIntegrate workflow. (Illustrative schematic, not sample data.)

Built for the questions you are asking

The platforms, pipelines and standards behind this service.

OmniIntegrate pipeline

Nucleome's multi-omics integration workflow: MOFA+/DIABLO-class latent-factor modelling, joint pathway analysis, correlation networks, and machine learning — with permutation-based validation, not over-fitted claims.

Neoantigen discovery pipeline

End-to-end tumour-antigen prioritisation: somatic variants, expression confirmation, HLA typing, and peptide–MHC binding prediction — candidates ranked with evidence tiers for downstream validation.

siRNA discovery pipeline

Rules-based siRNA design with off-target minimisation: seed-region specificity, isoform coverage, SNP/variant avoidance, and ranked candidate guides for empirical testing.

Versioned, reproducible workflows

Every analysis runs on pinned software versions and reference data — your results can be regenerated and defended.

Full omics coverage

Genomics, transcriptomics, epigenomics, metagenomics, proteomics, and metabolomics — one team for all layers.

Consultation-first model

Study design, statistical power, and analysis planning are part of the service — before any data is processed.

Data analysis in pictures

Illustrative analyses showing the outputs your project report contains. All data below are simulated for illustration — not client results.

-3-2-10123-2-1012Latent factor 1Latent factor 2Responders (n=40)Non-responders (n=40)
OmniIntegrate — latent-factor separation of responders. Latent-factor analysis of a simulated multi-omics cohort (transcriptome + proteome + metabolome + genotype) integrated with a MOFA+-class model. Responders and non-responders separate on the first two latent factors; factor loadings identify which omics features drive the separation. Permutation testing and cross-validation accompany every model we deliver. (Simulated data for illustration — not client results.)
11010010001000011010050 nM500 nM10 TPMN1N2N3N4N5N6N7N8N9N10N11N12Predicted pMHC binding, IC50 (nM, log scale; lower = stronger)Tumour expression (TPM, log scale)Tier 1Tier 2Tier 3
Neoantigen discovery — evidence-tiered candidate ranking. Neoantigen candidates from a simulated tumour (12 candidates) ranked by predicted pMHC binding strength (IC50) and tumour RNA expression. Tier 1 (strong binding, high expression) is prioritised for validation; Tier 3 is deprioritised. Every candidate is a computational prediction requiring empirical validation — the report states this explicitly. (Simulated data for illustration — not client results.)

Service highlights

What you can expect from our team in Hyderabad.

Analysis-only flexibility

Bring raw data from any source — your lab, a collaborator, or a public repository — or generate it with Nucleome.

OmniIntegrate pipeline

Multi-omics integration with validated models — not over-fitted claims.

Neoantigen & siRNA pipelines

End-to-end candidate discovery with evidence tiers and honest validation requirements.

Reproducible by default

Pinned versions, documented parameters, and full provenance in every deliverable.

Full omics coverage

Genomics, transcriptomics, epigenomics, metagenomics, proteomics, metabolomics — one team.

Consultation-first

Study design and statistical power advice before analysis begins.

What this service does and does not cover

Scope note. This is a data-analysis-only service: raw data must be supplied by the client or generated by Nucleome. Neoantigen and siRNA candidates are computational predictions requiring empirical validation — we tier evidence honestly and flag what remains unvalidated. Predictive models are reported with their validation status and limitations. Turnaround times and pricing are defined per project.

Frequently asked, in plain language

From the basics to the finer points of design and interpretation.

basicDo I need to be a sequencing client?
No — this is an analysis-only service. Provide raw data from any source, or generate it with Nucleome.
basicDo you analyse plant, animal and microbial data?
Yes. Our Agrigenomics division handles crop, livestock and microbial projects, including T2T and de novo genome assembly, microbial genome assembly, RNA-seq, population genomics for breeding, and agri-metagenomics.
basicWhat data formats do you accept?
Standard raw formats — FASTQ/BAM/CRAM for sequencing, VCF for variants, raw MS files for proteomics/metabolomics, and processed matrices where appropriate.
basicWhat is OmniIntegrate?
Nucleome's multi-omics integration pipeline — joint analysis of two or more omics layers using latent-factor modelling, joint pathway mapping, and validated machine-learning models.
intermediateHow does the neoantigen discovery pipeline work?
Somatic mutation calling from tumour WGS/WES → RNA-seq expression confirmation → HLA typing → peptide–MHC binding prediction → candidates ranked with evidence tiers. Predictions are clearly labelled as predictions requiring empirical validation.
intermediateHow does siRNA design minimise off-target effects?
Seed-region specificity analysis, transcript-isoform coverage, SNP/variant avoidance, and off-target scoring — with ranked candidate guides for empirical validation.
intermediateHow do you ensure reproducibility?
Pinned software versions, documented parameters, containerised workflows, and full provenance — your analysis can be regenerated and defended.
advancedCan you reanalyse public datasets (TCGA, GEO)?
Yes — with full provenance and documented retrieval, useful for validation cohorts and hypothesis generation.
advancedCan you build a custom pipeline for my team?
Yes — bespoke workflow development (Nextflow/Snakemake-class) with documentation and deployment support.
advancedHow do you handle machine-learning models honestly?
Proper cross-validation, external validation where possible, permutation testing, and explicit reporting of limitations — no data leakage, no overfitting.
advancedCan OmniIntegrate handle my specific combination of omics layers?
Yes — genomics, transcriptomics, epigenomics, proteomics, and metabolomics in any combination; the integration approach is matched to the layers and question at design consultation.

Key literature & resources

Selected sources behind the methods and reference values on this page. Figures labelled "schematic" or "illustrative" are generated for this page, not client data.

  1. Argelaguet R, et al. MOFA+: a statistical framework for disentangling and integrating multi-omics data. Genome Biology. 2020;21:111.
  2. Singh A, et al. DIABLO: an integrative approach for identifying key molecular drivers from multi-omics assays. Bioinformatics. 2019;35(17):3055–3062.
  3. Jurtz V, et al. NetMHCpan-4.0: improved peptide–MHC class I interaction predictions integrating eluted ligand and peptide binding affinity data. The Journal of Immunology. 2017;199(9):3360–3368.
  4. Hoof I, et al. NetMHCpan, a method for MHC class I binding prediction beyond humans. Immunogenetics. 2009;61(1):1–13.
  5. Reynolds A, et al. Rational siRNA design for RNA interference. Nature Biotechnology. 2004;22(3):326–330.
  6. Jackson AL, et al. Widespread siRNA "off-target" transcript silencing mediated by seed region sequence complementarity. RNA. 2006;12(7):1179–1187.
  7. Dobin A, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21.
  8. McKenna A, et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Research. 2010;20(9):1297–1303.