Transcriptomics
Bulk and single-cell RNA, lncRNA and miRNA, taken from counts to pathways without writing code.
A privacy-first, browser-native omics platform where every researcher runs single- and multi-omics analysis, no code, no fragmentation, no compromises.
Bulk and single-cell RNA, lncRNA and miRNA, taken from counts to pathways without writing code.
Olink and protein-level data, tied to the same pathway identifiers as every other layer.
Metabolites mapped onto pathways and reactions, so a changed compound points to a biological process.
Lipid species grouped by class and linked to the pathways they take part in.
Microbial features brought into the same analysis as the host, not a separate workflow.
Gene, protein, metabolite, lipid and microbe data meet in one set of pathway identifiers. No code, no fragmentation.
Hear from researchers using Om™ to accelerate discovery, unify analysis, and move from raw data to actionable insights.
Study design, analysis and interpretation for research institutions, hospitals and biotechnology companies.
Learn more arrow_forward schoolStructured training programmes, workshops and certification pathways in multi-omics data science.
Learn more arrow_forward forumQuestions about institutional deployments, data governance or a collaboration? Our team will reply.
Contact arrow_forwardJoin research institutions, biotech teams, and pharmaceutical organizations using Om™ to accelerate discovery.
Privacy by design. Researchers and institutions remain in full control of their data, with storage, retention, and sharing governed entirely by their own policies and preferences.
Om™ is the multi-omics analysis flagship platform of OmVitalis. It automates and connects data validation, preprocessing, annotation, statistical analysis, visualization, enrichment and cross-omics integration, with governed deterministic AI-assistance, all in a single unified computational environment. No coding and No Fragmentation.
The next frontier in tissue-resolved biology. SpatialOm™ integrates spatial transcriptomics, spatial metabolomics, and spatial proteomics into a unified analytical environment, revealing the molecular geography of disease with unprecedented resolution.
From variant calling to pharmacogenomics. PrecisionOm™ translates genomic complexity into clinical and research utility, GWAS, polygenic risk scoring, variant interpretation, and population-level genomics, purpose-built for the GCC's unique genetic landscape.
Biology changes over time, OmniOm™ captures it. A proprietary engine for cross-omics concordance analysis and temporal decoupling, revealing how molecular layers diverge, converge, and drive each other across time, treatment, or disease progression.
Each workflow is purpose-built for its omics layer, fully integrated with the Om™ unified analysis engine. Click any workflow to explore its complete pipeline and real outputs.
Every omics workflow follows one rule: single data entry, continuous unbroken flow, from raw measurements to publishable discovery.
“OmVitalis is a computational biology technology company building integrated, reproducible software infrastructure for multi-omics research. The name OmVitalis reflects our scientific vision. ‘Om’ represents omics and omni-scale integration, while ‘Vitalis’, Latin for ‘of life’, embodies our dedication to understanding biology across every molecular layer. Together, they symbolize our mission to transform the complexity of life into meaningful scientific, institutional, and clinical impact. Omni Insights from the Omics of Life.”
OmVitalis is structured around three complementary branches, each addressing a distinct need in the journey from raw data to real-world impact.
SpatialOm™ (spatial omics), PrecisionOm™ (precision genomics) and OmniOm™ (cross-omics concordance and temporal analysis) are in development.
OmVitalis Consultancy delivers expert-led scientific advisory across the full omics research lifecycle, from study design and data generation strategy through to multi-omics integration, biological interpretation, and publication-ready reporting.
Architect of the OmVitalis computational platforms and omics suites. Specializes in integrative metabolomics, multi-omics data interpretation, and translating complex omics datasets into actionable biological insight across infectious disease, cancer biology, cardiovascular research, and longevity science.
"OmVitalis builds multi-omics intelligence platforms that help organizations integrate, analyze, and interpret complex biological data. Through a unified computational platform, expert consulting, and structured training, we transform raw omics data into actionable insight that accelerates research, innovation, and translational impact."
OmVitalis Systems was founded on a single conviction: the most powerful computational biology tools should not be locked behind years of bioinformatics training, fragmented software ecosystems, or inaccessible codebases.
We build platforms that bring the full power of multi-omics science (metabolomics, lipidomics, transcriptomics, proteomics, metagenomics, single-cell, causal inference, and more) into a cohesive researcher-native omics workspace for single- and multi-omics discovery.
Our platforms serve academic researchers, biotechnology companies, pharmaceutical teams, and clinical research institutions, providing the same expert-level capabilities previously requiring a dedicated bioinformatics team.
Publication-grade statistical methods, deterministic workflows, and journal-ready outputs.
A privacy-first experience designed for all researchers, not only coders.
Led by peer-reviewed scientists with international academic training across integrative omics and translational research.
Expert in genomics and bioinformatics with a strong focus on gene regulation, genomic interpretation, and computational biology in complex disease research. Strengthens OmVitalis genomic and transcriptomics consulting capabilities across multi-layered biological interpretation projects.
Specialist in single-cell transcriptomics and immuno-genomics with expertise in high-resolution gene expression profiling and immune response variation. Contributes advanced capabilities in single-cell analysis, transcriptomic integration, and immune pathway interpretation.
Define research objectives, select appropriate omics layers, establish power, sample requirements, and analytical scope aligned with your scientific goals.
True cross-omics integration across genomics, transcriptomics, proteomics, metabolomics, lipidomics, and metagenomics, revealing system-level biological signals.
Expert-guided translation of omics outputs into pathway-level insight, mechanistic understanding, and clinically relevant biological narratives.
Multi-omics biomarker panel construction, ROC analysis, causal inference modeling, and Mendelian Randomization for rigorous translational evidence.
Publication-ready figure generation, statistical reporting, methods writing, and grant application scientific strategy aligned with high-impact journals.
Strategic advisory for research institutions, biotech, and innovation programs seeking to build or scale omics research infrastructure and analytical capacity.
OmVitalis Consultancy provides scientific consulting and advisory services. We do not provide clinical diagnostic or medical treatment services.
OmVitalis Academy delivers rigorous, hands-on omics training programs, from foundational bioinformatics to advanced multi-omics interpretation, designed for researchers, clinicians, and institutional teams working at the frontier of life sciences.
Short video lessons and interactive guides on analysing omics data, free for every researcher. Start with Multi-omics, Demystified, then watch the same steps run in Om™.
Intensive hands-on training covering a specific omics domain. Ideal for research teams seeking focused, practical upskilling.
Deep immersion programs with structured daily sessions, practical exercises, and real dataset case studies across multiple omics layers.
Multi-session structured programs with assignments, deliverables, and certificates of completion aligned with professional development needs.
Fully tailored programs designed around your datasets, workflows, team composition, and institutional research priorities. Online or in-person.
Connect biological signals across genomics, transcriptomics, proteomics, metabolomics, lipidomics, and epigenetics to build interpretable biological narratives and biomarker discovery pipelines.
Metabolomics interpretation from preprocessing to pathway insight, including lipid class profiling, acyl-chain analysis, and biomarker discovery workflows.
Structured introduction to bioinformatics pipelines and computational workflows required for modern single and multi-omics projects, no coding prerequisites required.
Practical statistics program designed specifically for biological and omics datasets, covering study design, hypothesis testing, and publication-grade statistical methods.
Specialized program in single-cell transcriptomics interpretation, immune profiling, cell-state discovery, trajectory inference, and RNA velocity analysis.
Strategy-focused training to help teams build fundable, publishable omics projects, from methods writing and figure standards to journal targeting and grant narrative design.
Protein-level discovery workflows covering preprocessing, differential abundance, pathway interpretation, biomarker prioritization, and clinical translation.
Microbiome training from raw reads to taxonomy, diversity, functional profiling, differential abundance, and host-microbiome interpretation.
Variant analysis, annotation, prioritization, clinical phenotype integration, reporting logic, and responsible interpretation of WGS datasets.
End-to-end RNA-seq training with specialized mRNA, lncRNA, and miRNA workflows, differential expression, enrichment, and regulatory interpretation.
ROC/AUC, feature selection, cross-validation, survival analysis, model interpretability, and biomarker panels for translational research.
DAG thinking, mediation analysis, Mendelian randomization concepts, causal diagrams, and mechanistic interpretation across omics layers.
OmVitalis partners with regional and international conferences to deliver expert-led omics workshops, hands-on training sessions, and scientific presentations. We welcome collaboration with conference organizers, scientific societies, and academic institutions.
Tell us about your institution and training needs, OmVitalis Academy will prepare a tailored proposal including modules, duration, delivery format, and learning outcomes.
Whether evaluating a license, planning a deployment, or curious about Om™'s capabilities, we respond within 1 business day.
Academic requests receive priority review. We respond within 1 business day.
From raw LC-MS/MS data to structural insight, lipid class profiling, acyl-chain enrichment, pathway and class enrichment using LIPID MAPS, with direct integration into genomics (lQTL) and multi-omics analysis.
Step-by-step video guides for every stage of the lipidomics workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
Comprehensive metabolomics analysis from raw peak tables or public datasets, from metabolite annotation through pathway enrichment, phenotypic correlations, causal inference, and mQTL integration. Supports untargeted and targeted LC-MS profiles.
Step-by-step video guides for every stage of the metabolomics workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
End-to-end bulk transcriptomics from raw count matrices, with dedicated biotype-specific workflows for lncRNA and miRNA, publication-standard differential expression, and pathway enrichment across GO, Reactome, and MSigDB. Integrates directly with eQTL and multi-omics analysis.
Six short walkthroughs of a bulk transcriptomics analysis in Om™, from preprocessing to pathway enrichment. Free on OmVitalis Academy.
info New tutorials premiere on Thursdays. See the full course on OmVitalis Academy or start exploring Om™ to follow along.
End-to-end Olink NPX proteomics, from format validation and LOD filtering through Limma EB differential analysis, protein network biology, drug target annotation, and pQTL integration with full cross-omics connectivity.
Step-by-step video guides for every stage of the proteomics workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
Complete scRNA-seq from raw count matrices, ambient RNA removal, QC, normalization, batch correction, dimensionality reduction, clustering, automated cell type annotation (CellTypist/Azimuth), RNA velocity, trajectory, cell-cell communication, and true multi-omics integration.
Step-by-step video guides for every stage of the single cell transcriptomics workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
Comprehensive microbiome profiling from 16S rRNA amplicon and whole-genome shotgun sequencing, taxonomic classification, alpha/beta diversity, differential abundance with LEfSe, HUMANn3 functional profiling, ARG resistome analysis, and full multi-omics integration.
Step-by-step video guides for every stage of the metagenomics workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
Connect up to 5 omics layers in a single session, cross-omics correlation networks, chord diagrams, dysregulation Sankey flow maps, integrated heatmaps, co-abundance modules, and multi-omics biomarker discovery with ROC analysis.
Step-by-step video guides for every stage of the multi-omics integration workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
A privacy-first platform offering all four integrative QTL types, mQTL, lQTL, eQTL, and pQTL, with genome-wide Manhattan plots, interactive locus browser with LD heatmaps, cross-omics QTL concordance, and GWAS integration (coming soon).
Step-by-step video guides for every stage of the genomics · qtl analysis workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
A privacy-first, GUI-driven causal inference engine for multi-omics data, DAG builder, Mendelian Randomization (IVW/MR-Egger/WM), multi-layer mediation analysis, What-If counterfactual simulator, and integrated multi-omics biomarker discovery with ROC analysis.
Step-by-step video guides for every stage of the causal inference & biomarker analysis workflow in Om™. New tutorials added regularly.
info More tutorials are being recorded and will be available shortly. Start exploring Om™ to follow along.
Privacy by design. Researchers and institutions remain in full control of their data, with storage, retention, access, and sharing governed entirely by their own policies and preferences.
OmVitalis is built around a privacy-first philosophy: researchers and institutions define how scientific data is stored, retained, accessed, shared, and governed.
Researchers and organizations determine how data is stored, retained, shared, accessed, and governed across the research lifecycle.
Data policies can be aligned with institutional requirements, study protocols, data-sharing agreements, and compliance expectations.
OmVitalis does not access, retain, reuse, sell, or share project data unless explicitly authorized by the researcher or institution.
OmVitalis empowers researchers and institutions with advanced omics capabilities while preserving full control over how scientific data is governed, stored, retained, and shared.
OmVitalis combines governance-first data handling with practical security controls for users, teams, and institutions.
Project access can be restricted to authorized users, teams, and institutional stakeholders according to the governance model selected for each deployment.
Account sessions can be protected through authentication, timeout protection, and secure access checks to reduce unauthorized use.
Payments are handled through Stripe-secured infrastructure. OmVitalis does not store payment card data inside the Om™ system.
Data handling expectations can be documented clearly so researchers, compliance officers, and institutions understand how data is governed.
OmVitalis supports conversations with institutional review boards, data protection teams, IT departments, clinical collaborators, and enterprise partners.
Data handling can be aligned with institutional policies for research data, clinical metadata, user permissions, and project retention.
OmVitalis can support the governance expectations of universities, hospitals, biotechnology companies, pharmaceutical teams, and public sector organizations.
Researchers and institutions remain the decision-makers for data access, storage, retention, sharing, and project-level permissions.
Om™ is registered intellectual property under UAE law, and a U.S. provisional patent application has been filed for its systems and methods.
Patent pending refers to a filed provisional application. No patent has been granted.
For research institutions and large teams requiring additional governance.
Integrate with your institution's identity provider. Centralized access control, automatic provisioning, and one-click onboarding for your team.
A dedicated support contact for your institution, with priority response times and direct access to the Om™ technical team for security reviews.
OmVitalis can evolve toward managed, institutional, hybrid, or enterprise deployments while preserving researcher and institutional control of data.
OmVitalis supports scientific analysis, biological interpretation, education, and research workflows. It is not intended to replace clinical judgment, regulatory review, validated diagnostic systems, or institutional compliance processes.
Researchers and institutions are responsible for ensuring that data use, consent, ethics approval, privacy requirements, and regulatory obligations are followed in their specific research setting.
Methods, statistics and definitions used inside Om, written so you can see what a result means and what stands behind it.
Five main human layers shown. One dot is 100 sets.
Om Library release 2026.10.10. Gene, protein, metabolite and lipid counts are sets, microbe is signatures. The total covers human and mouse.
The Om Library ships inside Om as its knowledge engine. Every release is dated and counted, the sources are open, and where we have checked the library against known biology the result is stated here, together with what we have not tested yet.
These are examples, not the full list. The library is built only from sources that permit commercial redistribution.
Every workflow opens with an import step matched to the data type: peak tables for metabolomics, LC-MS/MS lipid tables, raw count matrices for transcriptomics and Olink NPX files for proteomics. Automated checks cover sample and feature quality before any statistics are run.
Values below the limit of detection are handled explicitly rather than silently. Metabolomics uses LOD/2 imputation, lipidomics applies LOD filtering, and Olink data use assay-specific detection thresholds. Imputing half the detection limit keeps censored values small but non-zero, which suits log-scale analysis.
Ambient RNA is background transcript from lysed cells that is captured in droplets with intact cells. Doublets are droplets that contain two cells. Om removes ambient signal with SoupX or DecontX and flags doublets with DoubletFinder or Scrublet before clustering.
Features are matched against reference compound libraries. Each annotation carries a Metabolomics Standards Initiative confidence level from 1 (identified compound) to 4 (unknown), so the strength of an identification is always visible next to the result.
Om accepts 16S rRNA amplicon data (V3 to V4) and whole-genome shotgun data. Amplicon sequence variant tables from DADA2 and QIIME2 are supported for taxonomic profiling.
Metabolomics data are log2 transformed, median normalized and Pareto scaled. Lipidomics data use IQR normalization with log2 transformation. Transformation stabilises variance across the intensity range, and scaling controls how strongly high-abundance features dominate multivariate models.
Raw count matrices from bulk RNA-seq, lncRNA and miRNA can be normalized with TMM, CPM or voom. TMM corrects for differences in library composition between samples, and voom estimates the mean-variance relationship so that counts can be modelled with linear methods.
Technical and clinical covariates such as batch, sex and other clinical variables are modelled with mixed linear models, so differences between groups are estimated after accounting for known sources of variation.
Single-cell counts are normalized with Scran, and highly variable genes are selected before dimensionality reduction. Batch effects across samples can be corrected with Harmony, ComBat or scVI.
Limma fits a linear model to every feature and moderates the variance estimates with empirical Bayes shrinkage. This stabilises inference when sample sizes are small. Results are presented as volcano plots, heatmaps, box plots with significance brackets and ranked feature lists.
Principal component analysis with density overlays and confidence ellipses, k-means clustering and hierarchical heatmaps show structure in the data before group comparisons are made. Differential abundance testing then quantifies which metabolites differ between groups.
Differences are tested at lipid class level and at species level. A dedicated acyl-chain panel summarises chain length, degree of saturation and double bond position, so changes can be read as structural patterns and not only as individual lipids.
Pairwise correlations are computed between features, between features and phenotypes, and between phenotypes. Batch screening runs the same analysis across all phenotypes in the dataset.
Alpha diversity within a sample is reported with Shannon, Simpson, Chao1 and ACE indices. Beta diversity between samples uses Bray-Curtis and UniFrac distances, and group differences are tested with PERMANOVA.
LEfSe and ANCOM-BC identify taxa that differ between groups. Results are shown as cladograms with LDA scores and as effect size plots.
PCA, UMAP, t-SNE and PHATE embed cells in low dimensions, with an interactive 3D UMAP viewer. Leiden and Louvain clustering partition cells, and the resolution parameter can be tuned to the biological question.
Clusters are labelled automatically with CellTypist or Azimuth, then refined by manual curation using marker genes and per-cluster differential expression.
RNA velocity with scVelo and pseudotime trajectories with Monocle describe how cells progress between states. CellChat infers cell to cell communication, and MAST tests differential expression in single-cell data.
Over-representation analysis asks whether a pathway contains more significant features than expected by chance. It works on a list of significant features and is a good first view when a clear significance cutoff exists.
GSEA ranks all measured features by a test statistic and asks whether the members of a gene set concentrate near the top or bottom of that ranking. No significance cutoff is needed.
In addition to testing for over-representation, Om scores pathway activity quantitatively so that the direction and magnitude of change can be compared across pathways and linked to clinical covariates.
The same analysis runs across several pathway databases at once. Results are brought together in a multi-database Manhattan view, bubble plots and a consensus table, so findings that hold across databases are easy to separate from single-database results.
Library matches are graded by where they come from: a database, a curated source, an inference or a prediction. The grade stays attached to the result, so an inferred match is never read as a database fact and a prediction is never presented as a curated one.
Some compounds are missing from every pathway collection. For these, a class can be derived from the nomenclature of the compound itself, for example a lipid species name. Sulfate and glucuronide conjugates can also be traced to their parent compound. Both are labelled as inferred and are kept apart from database sets.
Where a reaction is known to run in one direction, the library records it. Direction evidence is graded in five levels, from fixed by a metabolic model to unknown, each with a confidence. A step without evidence is shown as unknown rather than guessed, and for most reversible reactions direction is still unknown.
Full layers are available for human and mouse, with ortholog mapping so that a selection can carry across species. Other species are not covered yet.
Up to five omics datasets are linked through shared sample identifiers. Om maps features automatically, detects overlap between datasets and reports sample concordance before any integrated analysis starts.
Chord diagrams show the direction and strength of associations between features from different omics layers. A network view adds feature filters and sign filtering.
Sankey diagrams trace how dysregulation flows across omics layers, and integrated heatmaps show patterns of co-regulation.
Multi-omics modules group features that change together across samples. Module trends and pathway enrichment connect each module to biology.
Protein-protein interaction networks are built from differential results. Community detection finds functional groups, hub proteins are identified, and drug target annotation is added.
Biomarker panels are built with logistic regression and evaluated with ROC curves and AUC. Cross-validated feature importance indicates which features carry the signal.
Estimates which transcription factors, kinases, phosphatases, miRNAs or enzymes are consistent with the observed change in their known targets. Results are worded "consistent with", report targets measured out of targets known, and are tested against two null models. The approach has been checked on constructed data and on a real cohort for coverage and null behaviour; validation on real perturbation experiments is still to come.
Site-level data can be read as kinase or phosphatase activity through the known substrates of each enzyme. A phosphatase is scored with the sign reversed, and the result says so. This has been tested on constructed data only.
Modified sites are grouped by modification type and by the enzyme known to act on them, so a site-level result can also be read at set level. Sites are recognised by their common names, for example AKT1_S473.
Links metabolite data to the enzymes that act on those compounds, and gene data to the compounds those genes act on, through reaction models for human and mouse. A scan that calls an unusually large share of its units is flagged as a broad response instead of being reported as a list of findings.
Non-coding genes, miRNA families and curated miRNA and target pairs support seed enrichment, an estimate of implied miRNA activity, and miRNA to target integration. Seed results are corrected for 3 prime UTR length and GC content, and both the corrected and uncorrected view can be inspected.
Beyond taxa, the library covers biosynthetic gene clusters and antimicrobial resistance genes, independent of host. Signature sets from published host-associated studies are available for human and mouse, and small samples are labelled as low power.
A quantitative trait locus is a genetic variant associated with a molecular trait. Om supports mQTL, lQTL, eQTL and pQTL analysis using linear regression with covariate adjustment, and presents results in an interactive Manhattan plot and a locus browser with linkage disequilibrium heatmaps and gene tracks.
QTL signals are compared across omics layers to find pleiotropic loci, where one variant influences several molecular phenotypes, and to describe shared genetic architecture.
Mendelian randomization uses genetic variants as instrumental variables to estimate causal effects. Om runs two-sample analysis with inverse-variance weighted, MR-Egger and weighted median estimators, together with heterogeneity tests and pleiotropy diagnostics.
A directed acyclic graph encodes causal assumptions. In the DAG builder, prior biological knowledge can be drawn directly onto the graph, and structure learning suggests additional edges from the data.
Mediation analysis separates a total effect into direct and indirect components. Om extends this across omics layers, with path diagrams, mediation heatmaps and enrichment of the mediators.
The what-if simulator models the expected biological consequence of an intervention, so hypotheses can be ranked before laboratory work begins.
Reads regulator effects through a pooled signed network, reporting direct and indirect targets separately. The sign multiplies along each path, so a path through an inhibitor reverses the expected effect. Depth is limited to three steps. Tested on constructed data and for null behaviour on a real cohort.
Normalized Protein eXpression, the relative unit reported by Olink assays. NPX is on a log2 scale, so a difference of one NPX corresponds to roughly a doubling of the protein level. It is a relative measure and not an absolute concentration.
The lowest signal that can be reliably distinguished from background. Values below the LOD are censored and need explicit handling.
The expected proportion of false positives among the results called significant. Benjamini-Hochberg adjustment controls the FDR, whereas Bonferroni correction controls the stricter family-wise error rate.
An exact sequence inferred from amplicon reads after error correction. ASVs replace the older practice of clustering reads into operational taxonomic units.
The non-random association of alleles at nearby genetic positions. The r squared statistic measures how strongly two variants are correlated, and it explains why many neighbouring variants share one association signal.
In Mendelian randomization, a genetic variant that is associated with the exposure, is independent of confounders and affects the outcome only through the exposure.
When a single genetic variant influences more than one trait. Horizontal pleiotropy can bias Mendelian randomization estimates, which is why Om reports diagnostics for it.
Any enrichable gene, protein, metabolite, lipid, site or microbe set. A count of sets is not a count of independent findings, because sets drawn from different collections overlap heavily.
A dated, content-hashed snapshot of the library. Om verifies every file on load and records the release with each analysis, so a result can be traced to the exact knowledge it used.
Entries summarise the methods named in each Om workflow. Exact parameters and settings are set inside the workflow itself.