Scientific workflow roadmap

Scientific coverage grows by completing family-specific evidence chains, not by counting modules. A workflow family becomes publicly stronger only when its inputs, scientific acceptance, execution posture, grounding, decision pressure, and experimental consequence survive the same bounded review.

Evidence progression

flowchart LR
    implementation["implemented scientific surface"] --> contract["typed workflow and acceptance contract"]
    contract --> assets["governed primary and challenge assets"]
    assets --> execution["identified Runtime lanes and artifacts"]
    execution --> grounding["support, contradiction, and context"]
    grounding --> decision["challenge, sensitivity, downgrade, refusal"]
    decision --> consequence["feasibility and observed outcome"]
    consequence --> posture{"public evidence burden met?"}
    posture -->|no| bounded["internal or review-grade support"]
    posture -->|yes| auditable["outsider-auditable bounded support"]

Every promotion remains bounded to its named corpus, run mode, comparison policy, biological context, and consequence evidence. A stronger scientific implementation does not automatically promote a family whose Runtime or downstream evidence remains weaker.

Current family baseline

Family	Declared product posture	Evidence already present	Primary work still required
DDA	review-grade bounded at the current black-box ceiling	MaxQuant review package, Comet/Sage challenge package, import-backed runtime custody, PSM/FDR/inference review	repository-owned raw search execution or equivalently strong live-engine evidence; authority surfaces must agree
DIA	outsider-auditable bounded	primary and matrix-shift packages, raw-executable review lanes, library-aware QC and comparisons	faithful-rerun consistency, chromatogram-level vendor parity, broader library and absent-peptide pressure
LFQ	review-grade bounded at the product level	cohort and sparse-contrast packages, raw-executable processing, missingness and normalization pressure	external truth beyond repeatability, broader cohort transfer, consistent product and Runtime authority language
multiplex	internal support only	TMTpro and channel-stress packages, raw-executable processing, interference surfaces	outsider review, stable transfer, laboratory consequence, and a defensible public decision packet
PTM	outsider-auditable bounded	localization and ambiguity packages, site mapping, raw-executable review	occupancy, function, regulatory consequence, and ambiguity-aware transfer evidence
targeted	outsider-auditable bounded	transition and carryover packages, targeted QC and follow-up records	vendor parity, calibration transfer, matrix interference, and broader assay consequence

The Workflow Families ledger and the generated release surfaces are the operative authorities. When they disagree, the narrower posture governs and the mismatch blocks stronger publication language.

Expansion priorities

Close current family evidence gaps

The first priority is consistency and depth for the six declared families:

• align workflow, black-box, rerun, grounding, and consequence authorities;
• replace import-only or report-conditioned evidence where a stronger execution claim is required;
• add transfer cases that differ materially in cohort, instrument, library, missingness, ambiguity, interference, or assay burden;
• preserve negative, refused, and inconclusive results as public evidence.

Promote implemented adjacent domains carefully

Core contains additional chemistry, format, annotation, proteoform, and workflow-building surfaces. Those capabilities remain component-level until a named family owns:

1. a scientific question and acceptance contract;
2. governed primary and challenge corpora with redistribution terms;
3. an executable or explicitly import-backed Runtime route;
4. comparison rules and known transfer boundaries;
5. grounding and contradiction pressure;
6. recommendation and laboratory consequence appropriate to the claim.

Glycopeptide-heavy analysis, broader spectral-library search, additional external-engine workflows, and wider proteoform interpretation should not be described as supported workflow families before that chain exists.

Promotion decision

Proposed change	Required response
algorithm or parser added without a family packet	document component scope; do not raise workflow posture
primary benchmark passes without a distinct challenge case	retain the existing posture and publish the missing transfer burden
Runtime completes but Core acceptance fails	publish execution evidence and scientific refusal separately
scientific and Runtime evidence pass but grounding is insufficient	keep the result descriptive; do not promote decision support
recommendation survives challenge but Lab refuses readiness	publish the refusal and retain advisory-only language
all required records pass under a named scope	update the workflow ledger and generated release evidence together

Evidence required for broader claims

A wider cross-family statement requires more than several independently strong families. It needs a declared common scope, comparable acceptance semantics, transfer evidence across materially different conditions, consistent Runtime posture, and downstream limits that remain valid across the combined set.

Until that evidence exists, public language stays family-specific. Inspect Current Capability Limits for active ceilings, Benchmark Assets for corpus governance, and the Release Readiness Matrix for blockers that prevent publication.