Open-source · Regulated · Reproducible

R infrastructure for rigorous data science

ReproStats builds open-source R packages and provides specialist consulting for teams that need their analyses to be reproducible, auditable, and regulation-ready — from clinical trials to pharma manufacturing.

# Audit your R analysis for reproducibility risks
library(reproducr)

r <- audit_script("analysis.R")
risk_score(r)

── reproducr risk report ────────────────
✓ changelog 0 high, 1 medium
✓ seed_check set.seed() found
⚠ locale_check sort() detected

# Certify outputs against a baseline
certify(list(model = fit, table = df))
✓ 2 outputs certified [SHA-256]

# Later — detect drift
check_drift(list(model = fit, table = df))
✓ model ok (delta = 0.0)
✓ table ok

Open-source packages

Tools for reproducible, regulated R

All packages are free and open source. Production-ready tools ship to CRAN; packages in active development are available from GitHub.

On CRAN
reproducr

Audit R scripts for reproducibility risks — breaking package changes, missing seeds, locale-sensitive operations. Certify outputs and detect drift across runs and environments.

Documentation
In development
regulog

Tamper-evident, hash-chained audit logging for R applications operating under 21 CFR Part 11 and EU Annex 11. Built for GxP Shiny deployments with full IQ/OQ/PQ validation documentation.

GitHub
In development
lineager

Row-level data provenance for dplyr pipelines. Tracks where every row in your analysis dataset came from and what happened to it — answering the audit question regulators actually ask.

GitHub
In development
estimandr

ICH E9(R1) estimand framework for clinical trials. Specify intercurrent event strategies, generate submission-ready estimand tables, and scaffold sensitivity analyses in a tidy R API.

GitHub

Expert R implementation for regulated environments

We help pharma, biotech, and clinical research teams deploy R safely in regulated contexts — from GxP Shiny applications to regulatory submission support.

Regulatory R implementation

Deploy R and Shiny in GxP-compliant environments. System validation, 21 CFR Part 11 audit logging, and user access controls built in from the start.

Reproducibility audits

Independent audit of R analysis pipelines against reproducibility risk criteria — package versions, stochastic functions, locale sensitivity, and output certification.

Package validation (IQ/OQ/PQ)

Validation documentation for R packages used in regulated contexts. Requirements traceability, qualification scripts, and summary reports ready for regulatory inspection.

Estimand specification

ICH E9(R1) estimand framework implementation for clinical trials. Pre-specification of intercurrent event strategies, sensitivity analyses, and submission-ready estimand tables.

SAS → R migration

Statistical equivalence validation for teams migrating from SAS to R. Formal equivalence testing across methods and formal documentation for regulatory submission.

Training

Workshops on R in regulated environments — reproducible workflows, package version management, audit logging, and regulatory submission best practices.

Who we work with

Built for teams where it matters

Our packages and consulting are designed for contexts where analytical results have real-world consequences.

Pharma & Biotech

Clinical statisticians

Reproducible clinical trial analyses, estimand specification, and regulated Shiny deployment.

Regulatory affairs

Submission teams

Package validation documentation, audit trails, and R-to-SAS equivalence for FDA/EMA submissions.

Research

Academic & CRO teams

Reproducible pipelines, output certification, and drift detection for long-running studies.


Why this exists

R is becoming the standard for statistical analysis in clinical trials and pharmaceutical research. But the infrastructure for using R in regulated environments — audit logging, reproducibility certification, estimand tooling — barely exists.

ReproStats was founded to close that gap. We build the open-source packages the R community needs for rigorous, regulation-ready work, and we offer the consulting expertise to implement them correctly in the environments where they matter most.

All our core packages are open source and free. We believe good scientific infrastructure should be a public good.

0.2
reproducr on CRAN
35+
breaking-change entries tracked
4
packages in the ecosystem
97%
test coverage

Ready to make your R analyses regulation-ready?

Whether you need a package, a validation document, or a full implementation — we can help.

Get in touch