[email protected]
Request a demo

LUPStatistical computing · CDISC · TLFs

One language from raw data to signed-off TLFs.

Galax LUP is a clinical statistical programming environment built on GLX, a block-based language designed for regulated analysis. Convert raw data to SDTM, derive analysis datasets, run PK and bioequivalence statistics, model, and render page-identical RTF and PDF outputs, all inside one audited, project-scoped workspace.

Request a demonstrationSee how it works →
  • Validated against reference results
  • Project-scoped, audited workspace
27
GLX blocks in the library
4
Editor views, one program
10
Guided, form-driven analyses
5
Output formats, RTF and PDF identical
  • CDISC SDTM
  • ADaM
  • Controlled terminology
  • 21 CFR Part 11
  • ALCOA+
  • FDA · EMA · ANVISA BE methods
  • RTF · PDF · HTML · TXT · CSV

How it works

Raw data to submission, in one traceable program.

Galax LUP takes source data from Galax EDC, lab and external vendors, and delivers SDTM, analysis datasets and TLFs for submission, medical writing and safety review.

  1. 01

    Source data

    Galax EDC, central and local labs, vendor transfers and legacy files.

    Raw filesProfiling
    OutputRaw
  2. 02

    Standardise

    SDA assistant drafts the source-to-SDTM map with per-row confidence; you approve it.

    GLX.SDTM
    OutputSDTM + SUPP--
  3. 03

    Derive

    Reusable derivations build analysis datasets with full traceability to SDTM.

    GLX.DERIVEGLX.TABLE
    OutputADaM
  4. 04

    Analyse

    Safety and efficacy tables, PK/NCA, bioequivalence and statistical models.

    GLX.AESGLX.PKSTATSGLX.MIXEDMOD
    OutputResults
  5. 05

    Report

    Titles, footnotes, spanning headers and pagination, page-identical in RTF and PDF.

    GLX.TLFREPORT
    OutputSealed TLFs

Capabilities

Everything a statistical programming team needs.

Write it as code, build it visually or follow a guided analysis: the result is the same audited GLX program.

01

GLX language & four editor views

Readable GLX blocks with syntax highlighting, in-editor help and a block library. Work in Code, Visual, Graphical node-canvas or Guided view; every view writes the same program.

02

SDTM conversion with an SDA assistant

Profile raw files, draft a source-to-SDTM mapping with a per-row confidence score, review and approve it, then run GLX.SDTM with controlled terminology, SUPP-- domains, study days, SV/SE derivation and record-level provenance.

03

Safety & efficacy TLF engines

Demographics, baseline characteristics, disposition, hierarchical AE/CM/MH tables, findings summaries with change from baseline, shift tables and frequency tables, with listings and figures from the same block.

04

PK, bioequivalence & modelling

Noncompartmental analysis to ADPP, BE statistics with ABE, RSABE, ABEL and NTI methods for FDA, EMA and ANVISA, dose proportionality, compartmental and population PK.

05

Statistical models

t tests, inference on proportions, ANOVA/ANCOVA, general linear models, linear regression with robust standard errors and stepwise selection, and mixed models for repeated measures.

06

Submission-quality reporting

GLX.TLFREPORT renders titles, footnotes, spanning headers, panels and pagination to HTML, RTF, PDF, TXT and CSV, with RTF and PDF of the same run held page-identical.

07

Therapeutic-area study library

Organise work as therapeutic area, indication, phase and study design, with programs, libraries, outputs, versions and an audit trail for each design. Clone and version designs.

08

Sealed, verifiable outputs

Bioequivalence exports are sealed with signer, timestamp and SHA-256/HMAC integrity, and any delivered file can be re-checked with Validate Output.

Four editor views

Write it your way. It is always the same audited program.

Switch views at any time. Every view reads and writes the same GLX source, so a guided analysis can be opened as code, reviewed and versioned like any other program.

Code

Readable GLX with syntax highlighting, in-editor help and the block library.

Visual

Form-based editing of every block option, with validation as you type.

Graphical

A node canvas that shows how datasets and blocks connect, end to end.

1234

Guided

Ten form-driven analyses for common questions, no code required.

GLX language

27 validated blocks. One consistent grammar.

Each block validates its inputs before it runs and writes datasets, tables, listings or figures. Block families are shown below.

Data handling & derivation6

GLX.SDTM
Raw data to SDTM, driven by mapping, terminology and SDTMIG metadata
GLX.DERIVE
Reusable derivation engine for analysis variables
GLX.TABLE
Dataset derivation with a braced, readable syntax
GLX.SORT
Sorting, de-duplication and keep/drop
GLX.PIVOT
Transpose with ID labels, copy variables and duplicate handling
GLX.APPEND
Vertical combination with variable standardisation

Figures & reporting2

GLX.CLINICALPLOT
Clinical figure engine for safety and efficacy graphics
GLX.TLFREPORT
Final rendering to HTML, RTF, PDF, TXT and CSV

Safety & efficacy tables8

GLX.DEMOGRAPHIC
Demographics with categorical and continuous statistics and figures
GLX.BASECHAR
Baseline characteristics, including compound-condition rows
GLX.DISPOSITION
Subject disposition with matched denominators
GLX.ADVERSEEVENT
Hierarchical SOC/PT tables with any depth, thresholds and sort rules
GLX.FINDINGS / SUMMARY
VS, LB, EG and PC summaries with change from baseline and profiles
GLX.SHIFT
Baseline to post-baseline shift tables
GLX.CROSSTAB
Frequency and cross-tabulation tables
GLX.STATSPROFILE
Statistical profile of analysis variables

Statistics & models6

GLX.TTEST
Paired and two-group t tests with hypothesis and method options
GLX.PROPTEST
Inference on proportions
GLX.ANOVA
ANOVA and ANCOVA
GLX.GLMMODEL
General linear models
GLX.REGMODEL
Linear regression with robust and cluster-robust errors, selection and diagnostics
GLX.MIXEDMOD
Linear mixed models and MMRM

Pharmacokinetics & bioequivalence5

GLX.PKPARAMETERS
Noncompartmental analysis: AUC methods, λz, partial AUCs, steady state, urine, ratios
GLX.PKSTATS
BE statistics for 2×2, replicate and parallel designs; ABE, RSABE, ABEL, NTI; power
GLX.PKDOSEPROP
Dose proportionality by the power model
GLX.PKMODEL
One-, two- and three-compartment models with multi-start fitting
GLX.POPPK
Population PK by nonlinear mixed effects (FOCE, FOCEI, Laplace)

See it in action

Galax LUP in under a minute.

VideoGalax LUP: product walkthrough, sample study
Built for
  • Statistical programmers
  • Biostatisticians
  • Pharmacokineticists
  • Medical writers
  • QC reviewers

Quality and control

Designed for regulated clinical work.

  • Audit trail, archive-not-delete

    Every project keeps an audit trail. Records are archived rather than deleted, and use of deprecated grammar is logged.

  • Project-scoped data

    Data and programs are held per client and project in libraries, so nothing crosses between studies.

  • Roles and client isolation

    Users, roles and account locking, plus a separate login URL and server for each client.

  • Verified against reference results

    Statistical engines are checked against established reference outputs. Regression results agree with the reference to every reported statistic.

Works with

Connected across the Galax platform.

All modules →
EDCGalax EDC & eCRFClean clinical data, from protocol to database lock.AIGalax Gen AIA generative AI platform specialised for clinical data work.GSRSGalax Safety Reporting SystemSafety reports generated directly from study data.LABGalax Lab DataLaboratory results, samples and lab safety in one controlled place.eCDTGalax eCDTExternal vendor data, received, checked, reconciled and mapped.

Talk to the team that builds it

See your own outputs built in Galax LUP.

Bring a sample table, listing or figure and we will show how Galax LUP produces it, from raw data to a page-identical RTF and PDF.

Request a demonstrationRead our CDISC notes →