Shedding Hub

Every published measurement of how much pathogen a person sheds, and when — standardized, versioned, and ready to model.

Browse datasets

Above: SARS-CoV-2 stool shedding for eight patients from woelfel2020virological. Every vertex is a measurement; points on the dashed line fell below the limit of quantification.

Recently

What’s new

studies curated by AI

A multi-agent pipeline now finds, extracts, and cross-checks studies, and a human curator reviews every one before it lands.

How the pipeline works

197

fitted shedding curves

Every analyte with enough data is fitted three ways — exponential, gamma, and gamma-shifted — covering 128 analytes. Each fit is plotted on its dataset page.

See the fits

pip

install shedding-hub

Load any study, fit rise-and-decay curves that handle non-detects properly, and simulate cohorts. Published on PyPI with full documentation.

Read the docs

At a glance

What’s in the repository

68133

Measurements

28274

Participants

Studies

Pathogens

Specimen types

Explore

Interactive data explorer

Plot time courses, compare studies, and inspect shedding dynamics without writing code.

Why this exists

Shedding data is scattered by default

How much pathogen a person sheds, and for how long, drives biomedical research, transmission models, wastewater surveillance, and public health policy. Yet the measurements sit in supplementary spreadsheets, in figures and tables with various formats, and in lab methods and units that differ from one paper to the next.

Shedding Hub puts them in one schema, under version control, with the source paper attached to every study. Nothing here is a number you have to take on faith.

How it gets curated

Curated by people, scaled by machines

From May 2024, human biocurators extracted every study by hand, building the 39-study benchmark the rest of the pipeline is measured against.

Since June 2025, large language models do the first pass on discovery and extraction, and a second model checks the first. Expert review still gates what gets published, so scale never comes at the cost of the benchmark.

Read how the curation pipeline works →

Get started

From install to first plot in a few lines

Read the documentation

Community

Using Shedding Hub?

Tell us what you built

We want to know which studies you pulled, what you modeled, and what the data could not answer. It shapes what gets curated next, and we would like to credit the work on this site.

Share your use case

Get release notes

An occasional email when a batch of datasets lands, the schema changes, or the package ships something that affects your analysis. No other mail.

Subscribe to updates

How to cite

Citing the repository in a paper? Use this, and tell us so we can link it here.

Shedding Hub: curated biomarker shedding data. https://shedding-hub.github.io

Individual datasets should also cite their source study, linked from every dataset page.

Contribute

Have shedding data?

We take data from a single participant or an entire cohort, published or not. Our curators handle the schema work and the quality review, and contributors are credited on the dataset. If a study of yours is already in here and the extraction looks wrong, we especially want to hear that.

Contribute a dataset Report a data issue

People

Leadership

Meet the full team, including our students and curators →

Funding

The Shedding Hub was made possible by the Insight Net cooperative agreement CDC-RFA-FT-23-0069 from the CDC’s Center for Forecasting and Outbreak Analytics. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention. Support for the Shedding Hub is provided by the Emory Center for Infectious Disease Modeling and Analytics & Training Hub (CIDMATH).