Nexora · Literature-grounded workflow planning
Turn the field's consensus into an analytical plan
Nexora reads the methodology consensus from your Lumen session and produces an ordered analytical workflow plan where every parameter is justified by a citation, and where the literature is divided, it is flagged for you to decide rather than silently defaulted. Nexora plans; it does not execute.
runs a pipeline you already specified.
scaffolds best-practice steps from convention.
produces a plan where every parameter links to the published study supporting it, and flags where the literature is divided, before any analysis runs.
The methodology consensus to ground parameters in.
Pipeline type (or auto-detect from the data manifest).
Grounds the plan in your validated data.
Design context for the plan.
Picks the pipeline when Domain is Auto, and tailors the AI rationale.
Comma-separated site names. The plan becomes the harmonised baseline; per-site deviations are documented.
Understanding Nexora
Nexora turns the published literature's methodology consensus into a concrete, ordered analysis plan, where every parameter is justified by a citation, and where the field is divided, the choice is handed back to you. It plans the analysis; it does not run it.
What Nexora is
Choosing the parameters of a bioinformatics pipeline, trim lengths, references, tools, thresholds, is where reproducibility quietly breaks. Defaults get copied, choices go undocumented, and reviewers can't tell why a value was used. Nexora makes every choice explicit and traceable: it reads what comparable published studies actually did, and assembles a plan in which each parameter carries the evidence behind it.
It sits in the closed loop after Lumen (the literature), Praxis (the design) and Veraxa (the validated data), and produces the plan that Novora executes and Oris reports.
Why it's different
Runs a pipeline you already specified.
Scaffolds best-practice steps from convention.
Produces a plan where every parameter links to the published study supporting it, and flags where the literature is divided, before anything runs.
How a plan is built
- 1 · Start from a reviewable template
Each domain has a fixed, peer-reviewable pipeline structure. Tools are fixed by the template; only parameter values are decided from the literature.
- 2 · Read the consensus
For each parameter, Nexora draws on the methodology consensus your Lumen session extracted, the values comparable studies reported, with their citations.
- 3 · Arbitrate honestly
It decides whether the literature actually supports a value, supports it weakly, or is genuinely divided, and labels each parameter accordingly.
- 4 · Pin and export
Reference and tool versions are pinned for reproducibility, and the plan exports to the workflow format you use (and a manuscript methods section).
It defers, it doesn't guess
Every parameter ends in one of three honest states: grounded in the literature(with citations and a confidence level), a documented default (when the evidence is thin), or a decision for you (when comparable studies genuinely disagree). A single source is never called "high confidence," and a value too weakly supported is handed back rather than asserted. Nexora will not manufacture a consensus the literature doesn't support.
What you get
- An ordered plan with every parameter labelled, cited, and confidence-rated.
- A clear list of the choices the literature leaves to you, with the competing options and their support.
- Reproducible exports, Nextflow, Snakemake, CWL, JSON, plus a manuscript methods section.
- A citation bibliography (BibTeX / RIS / CSL) for every paper the plan rests on.
Supported domains
More domains are added over time as new reviewable templates.