bijux-proteomics-core

bijux-proteomics-core owns durable scientific contracts in bijux-proteomics. If foundation is the shared language, core is the package where that language becomes scientific law: workflow families, benchmark acceptance, lifecycle transitions, and runtime-agnostic workflow contracts are defined here before downstream packages are allowed to interpret, execute, or operationalize them.

That role is broader and more concrete now than older docs admitted. Core is not only a workflow-rule package. It is the repository's deepest scientific owner: sequences, chemistry, spectra and mzML intake, identification, quantification, PTM-facing review surfaces, benchmark asset packaging, and workflow contracts all converge here before downstream packages score or act.

flowchart LR
    foundation["foundation<br/>shared payload meaning"]
    core["core<br/>scientific law and benchmark contracts"]
    knowledge["knowledge<br/>evidence state"]
    intelligence["intelligence<br/>decision policy"]
    runtime["runtime<br/>execution control"]
    lab["lab<br/>assay loop"]

    foundation --> core
    core --> knowledge
    core --> intelligence
    core --> runtime
    core --> lab

Why This Package Exists

• the system needs one place where scientific law stays durable before runtime behavior or recommendation policy begins
• downstream packages should inherit benchmark-acceptance and lifecycle rules, not reinvent them
• public workflow-family trust language needs one contract owner that can be inspected independently of judgment or execution

What It Owns

• scientific law and benchmark-acceptance boundaries
• runtime-agnostic workflow contracts and lifecycle transitions
• benchmark package shapes, challenge-corpus seams, and flagship scientific contract language
• typed sequence, chemistry, spectra, mzML, identification, quantification, PTM, DIA, and review surfaces that downstream packages consume rather than recreate

Concrete Scientific Families

owner band	visible package substance	why it matters to readers
sequence and study	domain, sequences, study	sample, sequence, and design semantics are normalized before downstream interpretation
chemistry	chemistry, isotope_labeling, proteoforms	mass, fragments, isotopes, and modifications are first-class scientific owners
proteomics I/O	io, identification, interfaces	spectra, mzML, search normalization, and reviewable ingestion stop being hidden support glue
quantitative and review surfaces	quantification, review, ptm, targeted, dia, lab/qc	analytical and assay-facing outputs start from durable scientific contracts
benchmark and workflow law	benchmarks, workflow, interpretation	public workflow-family claims now have explicit evidence and contract roots

Why This Package Feels Heavier Now

• benchmark-backed public workflow evidence starts here instead of being scattered across reports
• runtime, knowledge, intelligence, and lab all inherit scientific structure from this package before adding owner-local meaning
• family-level trust pages depend on core package breadth more than any other owner surface in the repository
• deeper biology and chemistry breadth is now visible as real package-owned structures rather than inferred from later prose

Shared Reader Routes

• Use Benchmark Assets when the question is about public benchmark packages, lineage, or flagship acceptance before it becomes a package-local question.
• Use Workflow Families when the question is which family currently deserves attention.
• Use Decision Support when the question is whether a strong benchmark packet still loses later to grounding, judgment, or consequence.

Start Inside This Package

• Open Foundation for the package role, ownership boundary, and benchmark-facing proof chain.
• Open Package Overview when the question is which scientific families actually live here and why the package is heavier than it first appears.
• Open Flagship Public Benchmark Catalog when the question is which paired public benchmark packets the repository really ships.
• Open Architecture when the issue is internal structure rather than public benchmark evidence.
• Open Interfaces when the issue is a public contract or package-facing surface.

Reader Questions This Package Can Answer

• which workflow states, lifecycle transitions, and benchmark-acceptance boundaries are treated as program law
• where a workflow-family scientific claim becomes a durable contract instead of a provisional interpretation
• which lower-level biological and chemistry inputs are normalized here before downstream packages are allowed to reason over them
• whether a benchmark package reflects a real review surface or only a shallow export of convenience

What A Skeptical Reader Should Verify

• whether a claimed scientific family is backed by typed owner surfaces rather than by report prose alone
• whether chemistry, spectra, search, quantification, PTM, and targeted analysis are truly package-owned instead of being spread through demos and runtime wrappers
• whether benchmark packages point back to real scientific law inside core instead of acting as decorative fixtures

What It Refuses

• shared serialization primitives that belong in foundation
• evidence truth and contradiction policy that belong in knowledge
• recommendation posture that belongs in intelligence
• execution orchestration that belongs in runtime
• assay-worth and outcome meaning that belong in lab

Strongest First Proof

• start at Flagship Public Benchmark Catalog when the question is whether the repository has real public scientific packets
• continue to Benchmark Assets when the question becomes freshness, licensing, lineage, and acceptance
• hand off to runtime, knowledge, intelligence, or lab only after the core scientific contract is clear

First Proof Check

• packages/bijux-proteomics-core/src/bijux_proteomics
• packages/bijux-proteomics-core/tests
• flagship benchmark package manifests and acceptance-facing evidence routes
• architecture and interface pages once the question narrows to one local scientific contract