Skip to Content
Orphic
/

Data Analysis Expert

What it does

Two entry points, one pipeline: if you only have a research question, it finds runnable public datasets (search, cohort checks, study-design calls); if you already have a table or omics matrix, it goes straight into QC and analysis. Both paths converge on the same delivery discipline: differential expression, the depth ladder (result diagnostics + ranked-list enrichment assessment), a machine-readable output contract, and a literature-cross-checked report.

Typical flow (Line B: public data)

  1. Lane assessment — split the question into Q1..Qn, search public omics repositories, verify cohort/population match, and stop for you to confirm the chosen accession.
  2. Fetch and differential — pull the matrix, decide paired vs unpaired, input QC, differential expression, annotation and enrichment.
  3. Contract and write-up — produce analysis_manifest.json, pass the contract self-check, add supporting literature, and write up the report.

Typical flow (Line A: in-house data)

  1. Clarify and QC — ask the analysis question, data type, and grouping in one go; validate and snapshot-freeze the data.
  2. Analyze — route by data type: quick/formal/paired differential, batch correction, merging, single-cell and spatial; the depth ladder applies the same way (diagnostics + enrichment assessment are mandatory).
  3. Report and export — conclusions report, delivery audit, and export of results (with provenance in a sidecar file).

What it delivers

  • An analysis report: methods, statistics, conclusions and limitations, strictly separating “what I computed” from “what the literature says”;
  • Figures (differential, enrichment) and data files (matrices, DE tables) — downloadable, storable in the asset library;
  • A machine-readable output contract (analysis_manifest.json).

Capabilities used

Public dataset search and fetch, table validation/query/merge/snapshot/export, differential expression (quick/formal/paired), batch correction, result diagnostics and ranked-list enrichment, annotation and interpretation, contract self-check, literature cross-check, write-up and delivery audit.