A quantitative framework to assess clinical data readiness for regulatory submission.

R4SUB answers a single question in a reproducible, regulator-aligned way:

Is this clinical data package ready for regulatory submission?

The framework transforms fragmented validation outputs, metadata checks, and analysis diagnostics into a unified, explainable, quantitative readiness signal — the Submission Confidence Index (SCI).

R4SUB is not a replacement for validation tools. It is a readiness layer that builds on their outputs.

9 modular R packages — each independently testable, documented, and published.

Package	Purpose	CRAN	Docs
r4subcore	Evidence schema, parsers, scoring primitives		site
r4subtrace	ADaM↔SDTM traceability engine		site
r4subscore	SCI scoring engine		site
r4subrisk	FMEA-based risk quantification		site
r4subprofile	Regulatory authority profiles (FDA, EMA, PMDA, MHRA, HC, TGA)		site
r4subusability	Reviewer usability indicators		site
r4subdata	Synthetic example datasets		site
r4subui	Interactive Shiny dashboard		site
r4sub	Meta-package — one install loads the full ecosystem		site

install.packages("r4sub")   # installs and attaches the full ecosystem

R4SUB measures submission readiness across four orthogonal dimensions:

The SCI is a weighted composite score (0–100) across all four pillars, calibrated per regulatory authority.

The SCI is fully decomposable — every score traces back to concrete evidence rows.

r4subprofile calibrates SCI weights and required indicators per authority:

Clinical Data Assets (SDTM, ADaM, TLFs, Define.xml)
                         │
                         ▼
              ┌──────────────────────┐
              │     r4subcore        │  Evidence schema + parsers
              └──────────┬───────────┘
          ┌──────────────┼──────────────┬──────────────┐
          ▼              ▼              ▼              ▼
    r4subtrace      r4subrisk     r4subscore    r4subusability
    Traceability    Risk FMEA     SCI Engine    Usability checks
          └──────────────┴──────────────┴──────────────┘
                         │
                         ▼
              ┌──────────────────────┐
              │    r4subprofile      │  Authority calibration
              └──────────┬───────────┘
                         │
                         ▼
              ┌──────────────────────┐
              │      r4subui         │  Shiny dashboard
              └──────────────────────┘

library(r4sub)

# Load example data
data(evidence_pharma)          # from r4subdata

# Score
scores  <- compute_indicator_scores(evidence_pharma)
pillars <- compute_pillar_scores(evidence_pharma)
sci     <- compute_sci(pillars)
sci$SCI   # 0–100
sci$band  # "ready" / "minor_gaps" / "conditional" / "high_risk"

# Apply an authority profile
prof <- submission_profile("FDA", "NDA")
val  <- validate_against_profile(evidence_pharma, prof)
val$is_compliant
val$missing_indicators

# Launch the dashboard
r4sub_app(evidence = evidence_pharma)

Clinical Programmers · Biostatisticians · Regulatory Data Standards Teams · Quality Assurance · Submission Operations

We welcome contributions:

• New readiness indicators and scoring rules
• Additional regulatory authority profiles
• Traceability parsers for new source formats
• End-to-end workflow examples and demos
• Bug reports and feature requests

Open an issue or discussion in the relevant repository.

R4SUB — Because compliance is not the same as readiness.