Industries · Precision Medicine
Precision Medicine
Multi-omics characterisation for patient stratification, treatment response and biomarker validation.
Industries · Precision Medicine
Multi-omics characterisation for patient stratification, treatment response and biomarker validation.
Precision medicine depends on identifying biological differences between patients that matter for a defined clinical outcome. GenXMap integrates genomic, transcriptomic, proteomic and clinical data to identify molecular subtypes, associate them with clinical endpoints, and test whether candidate biomarkers hold beyond the cohort in which they were discovered. The objective is not to find differences between patients; it is to determine whether those differences define biologically coherent groups with measurable clinical relevance.
A designed programme rather than an analysis order: the strategy is set by the cohort, the biological question and the clinical endpoint, not by a predefined technology stack.
The work starts before the bench: how several data sources become one dataset before any of it is analysed — the step that decides whether a subtype is real or an artefact of how the data were merged. Decisions a protocol run to order never reaches.
Variant calling, differential expression and differential protein abundance per sample and group, with QC and exclusions recorded.
Bulk RNA-seq shows what changes across the population; single-cell shows which cell populations carry the change; spatial shows where it occurs in the tissue. They are added when the question requires them, not by default.
Unsupervised and consensus clustering across integrated layers; subtypes defined by molecular profile, not by outcome.
Each subtype tested for stability of patient assignment, molecular coherence, separation between groups and reproducibility across samples or cohorts, to distinguish a biological subgroup from a pattern created by sampling, technical variation or the discovery dataset.
Subtypes and signatures tested against the endpoint defined in the study protocol: response, progression, survival or another predefined outcome.
Signatures, classification criteria and model thresholds are locked before independent data are touched. Where an independent cohort exists, the signature is tested on data not used in its discovery. A signature that performs well only on its discovery cohort is reported as a candidate, not a biomarker.
Classification and survival models linking subtype to outcome, with parameters, thresholds and versions recorded per study.
RNA does not capture the complete molecular state of a patient. When a transcriptomic change and a proteomic change converge on the same pathway, the evidence is stronger than either alone; when a genomic alteration is reflected in a downstream expression programme, the subtype has a mechanism. Layers are integrated before interpretation rather than analysed as disconnected datasets. For response and resistance signatures in immuno-oncology specifically, see the Predictive Biomarker Discovery & Validation Programme.
GenXMap does not need to generate every dataset. Existing genomic, transcriptomic, proteomic, microbiome or clinical data, from previous studies, other sequencing providers, historical cohorts or public and collaborative sources, can be analysed on their own or integrated with newly generated data. What determines the analysis is the question, the study design and the metadata, not who generated the data.
Stratification is only as strong as the cohort that defines it. GenXMap supports study planning around discovery and validation cohorts, inclusion criteria, sample collection and longitudinal timepoints, clinical endpoints, metadata requirements and the statistical strategy, and can work with clinical, academic and industry partners where samples or additional cohorts need to be sourced.
To scope a programme we need three things: the cohort (size, sample types, existing data), the endpoint, and the decision the result will support.
The laboratory work behind these analyses runs on OMICS4, GenXMap’s platform site. Sample requirements, shipping and turnaround are published there.
Book the platforms on OMICS4Groups, endpoints, available metadata and the analytical strategy chosen.
Genomic, transcriptomic and proteomic results, with QC and exclusions.
Molecular clusters, classification criteria and stability assessment.
Candidate signatures, statistical results and association with predefined endpoints.
Performance on held-out or independent cohorts where available.
Written against the biological question the study started from, stating what the data support and what they do not.
Methods, software versions, parameters, thresholds, processed matrices and versioned pipelines, delivered for review, audit or reproduction by a third party
A transcriptomic or multi-omic signature that holds on the discovery set and fails on the next one has cost you a programme, not a study. Validation needs cohorts, clinical metadata and collaborators, and those are not things a sequencing provider has.
GenXMap is joining the Paris-Saclay Cancer Cluster, and is joining the Marseille Immunology Biocluster. It places us alongside academic groups, hospitals and biotechs, and lets us coordinate collection directly with the clinical and research teams working there. Where the route is transcriptomic, bulk, single-cell and spatial can be contracted end to end as our 360° transcriptomics offering, so a signature can be traced to a cell type and a location rather than a gene list.
Case studies for this industry are in preparation: get in touch to discuss a project directly.
Have a project in this space?