J-EXPRESS / OPEN SCIENTIFIC METHODS
Scientific skills.
Open to inspection.
Modern AI brings broad reasoning capabilities. Our research-informed skills add explicit methodological procedures, quality checks, limitations and references. J-Express tools turn those procedures into reproducible computation, while the researcher contributes context, judgment and domain expertise.
The most valuable insight comes from the combination: the AI's reasoning, evidence-based workflow skills, reliable analysis tools, and — crucially — the scientist's own experience.
Matrix analysisPCA, clustering, correspondence analysis and numerical matrix explorationNormalizationModality-aware transformations, scaling and preprocessing safeguardsEmbeddingsOriginal-space neighbors, UMAP, semantic search and representation analysisBulk RNA-seqCount-aware QC, experimental design and exploratory expression analysisSingle-cell RNA-seqCell QC, feature selection, neighborhoods and clustering with donor-aware caveats