Clinical Submission Readiness, Quantified

An open-source R ecosystem for evidence-based assessment of regulatory submission quality. Transparent, auditable, and built for the modern clinical data scientist.

# Assess submission readiness
library(r4subcore)

ctx <- r4sub_run_context(
  study_id = "STUDY001",
  environment = "PROD"
)

evidence <- p21_to_evidence(
  p21_results, ctx
)

Why R4SUB?

Submission readiness is more than validation reports. It's an evidence graph across datasets, metadata, traceability, and reviewer experience.

📊

Evidence-Based Scoring

Normalize heterogeneous sources (Pinnacle21, OpenCDISC, internal engines) into a unified evidence table. Transparent, reproducible scoring primitives.

🔍

Transparent & Auditable

No black-box algorithms. Every weight, threshold, and aggregation method is configurable and documented. Full traceability from raw input to final score.

🧩

Modular Ecosystem

Core contracts separate from domain logic. Mix and match packages for quality, traceability, risk, and usability assessment.

🛠️

Tool-Agnostic

Built to integrate with your existing validation infrastructure. Parsers for common tools, extensible for proprietary systems.

The R4SUB Ecosystem

Modular packages for comprehensive submission assessment

Available

r4subcore

v0.1.0

Foundational contracts: evidence schema, run context, parsers, scoring primitives. The boring, stable core.

  • Evidence table schema + validators
  • Pinnacle21 parser
  • Severity mapping & normalization
  • Indicator registry
Planned

r4subtrace

Roadmap: Q2 2026

Traceability analysis: ADaM spec ↔ code ↔ outputs. Derivation completeness, lineage graphs, coverage metrics.

  • Spec-to-output traceability
  • Code-to-dataset lineage
  • Derivation coverage scoring
Planned

r4subrisk

Roadmap: Q3 2026

Risk assessment: FMEA-style impact scoring, defect severity weighting, submission delay prediction models.

  • Defect impact classification
  • Risk heat maps
  • Predictive delay models
Planned

r4subscore

Roadmap: Q4 2026

Submission Confidence Index (SCI): multi-domain aggregation, calibration against real submissions, industry benchmarks.

  • SCI calculation engine
  • Multi-domain aggregation
  • Benchmark comparisons

Get Started

Install r4subcore and start assessing your submission in minutes

1. Installation

# Install from GitHub (CRAN submission planned)
# install.packages("pak")
pak::pak("R4SUB/r4subcore")

2. Create a Run Context

library(r4subcore)

ctx <- r4sub_run_context(
  study_id = "STUDY001",
  environment = "PROD"
)

ctx$run_id  # Unique identifier for this assessment

3. Ingest Validation Results

# Parse Pinnacle21 output
p21_data <- read.csv("pinnacle21_report.csv")

evidence <- p21_to_evidence(
  p21_data,
  ctx = ctx,
  source_version = "P21-3.1"
)

# View summary
evidence_summary(evidence)

4. Score and Analyze

# Aggregate by indicator
scores <- aggregate_indicator_score(
  evidence,
  by = "indicator_id",
  method = "weighted"
)

# View worst performers
head(scores[order(scores$score), ])

Documentation