yassa_bio is an open-source, validation-ready Python toolkit for bioanalytical assay data, starting with ELISA and other ligand-binding assays (LBA), with a roadmap toward universal support for plate-based and chromatographic assays.

It is built for scientific rigor, automation, and regulatory alignment—targeting labs that need reliable, explainable, and auditable pipelines, from academic cores to pharma and CROs.

• ✅ Curve fitting and back-calculation: Supports 4PL, 5PL, and linear models with transform, weighting, and outlier handling
• ✅ ICH M10 / FDA BMV 2018 / USP ⟨1225⟩-based acceptance criteria
• ✅ Automated evaluation of:
  • Calibration fit
  • QC accuracy/precision
  • Carryover
  • Selectivity, specificity
  • Dilution linearity
  • Stability
• ✅ Flexible schema for validation and analytical runs using pydantic models
• ✅ Extensible plug-in registry: Add new evaluators or transform logic via decorators
• ✅ Rerun-aware pipeline engine: Smart re-evaluation when calibration levels fail and are dropped
• ✅ Readable outputs: Results returned as dicts for use in HTML/PDF reports (planned)

This repo is in early alpha. Expect breaking changes.

While most modules are unit-tested for correctness, the full pipeline has not yet been validated against real-world or external datasets. Users should treat results as provisional and subject to change.

In addition:

• ❌ No graphical reports or CLI UX yet—data must be loaded programmatically
• ❌ No built-in loader UI: You must manually define PlateLayout and WellTemplate objects

This library is:

• Schema-first: All input/output structures are formalized via pydantic, enabling structured config, validation, and API wrapping
• Composable: Pipelines consist of atomic steps with a typed shared Context, rerunnable after refitting
• Domain-aligned: Acceptance steps follow bioanalytical validation standards like ICH M10, FDA BMV 2018, and USP ⟨1225⟩
• Deployable: Future enterprise modules (auth, signoff workflows, dashboards) will wrap this open core

from yassa_bio.run import run
from yassa_bio.schema.layout.batch import BatchData
from yassa_bio.schema.analysis.config import LBAAnalysisConfig
from yassa_bio.schema.acceptance.validation.spec import LBAValidationAcceptanceCriteria

ctx = run(
    batch_data=BatchData(...),
    analysis_config=LBAAnalysisConfig(),
    acceptance_criteria=LBAValidationAcceptanceCriteria(),
)

print(ctx.acceptance_results)

See PlateData and WellTemplate for how to define input formats.

git clone https://github.com/your-username/yassa_bio.git
cd yassa_bio

pipx install poetry         # or: brew install poetry
poetry install              # install dependencies
poetry run pre-commit install
poetry run pytest           # run tests

yassa_bio/
├── pipeline/         ← Core engine: step, pipeline, rerun, dispatch
├── schema/           ← Pydantic models: layout, acceptance, config
├── evaluation/       ← Step logic for acceptance criteria
├── core/             ← Curve fitting, transforms, QC utilities
├── io/               ← Raw data loaders (CSV, Excel)
└── run.py            ← Main entrypoint to run the pipeline

This is an early-stage open science tool. If you're in pharma, biotech, or academia and would like to collaborate, file an issue or email the maintainer. You can also:

• Fork the repo and open a PR
• Request features or improvements
• Share example plate formats or data

Apache 2.0 – free for academic, commercial, and clinical use.

Contact [ruba.andrew@gmail.com] or open a GitHub Issue.