Skip to content
/ KWELA Public

Specialized RT-QuIC classification for complex environmental matrices

Notifications You must be signed in to change notification settings

RFeissIV/KWELA

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 

Repository files navigation

KWELA

Kinetic Weighted Evaluation for Latent Amyloid

Specialized RT-QuIC classification for complex environmental matrices.

Overview

KWELA extends RT-QuIC (Real-Time Quaking-Induced Conversion) statistical analysis to complex environmental matrices through matrix-specific calibration. It provides the statistical framework needed to reliably apply RT-QuIC in environmental surveillance contexts where standard statistical approaches may be confounded by matrix-induced artifacts.

KWELA is a specialized complement to standard statistical methods (ANOVA, t-tests, AUC), not a replacement. Standard methods work well for laboratory conditions with clean samples. KWELA addresses the gap when environmental matrix complexity introduces artifacts that confound interpretation.

Status

CRAN submission pending. Package source will be available here upon acceptance.

Features

  • Robust Classification: Positive-control-anchored Z-score transformations with logistic calibration and kinetic shape gating
  • Auto-Detection: Automatically detects RT-QuIC vs Nano-QuIC and applies optimized thresholds
  • Interference Detection: Continuous interference score based on metric correlation dissociation
  • Concordance Analysis: Per-treatment CCC/RMSE/MBE against positive control profiles
  • Predictive Projection: Projects classification reliability based on metric-space distance from positive controls
  • Classification Entropy: Measures boundary quality per treatment group
  • Robust Statistics: Adaptive scale estimation (MAD, Sn, Qn) with Rousseeuw-Croux (1993) finite-sample corrections
  • Plugin Architecture: Accepts standard data frames from any upstream source (e.g., quicR)

Installation

# Available after CRAN acceptance:
# install.packages("KWELA")

Quick Start

library(KWELA)

# Run the built-in demo
demo_kwela()

# Analyze your data
result <- kwela("my_rtquic_data.csv")

# With full evaluation (interference, concordance, projection)
result <- kwela(my_dataframe, evaluate = TRUE)
eval_report <- attr(result, "evaluation")

When to Use KWELA vs Standard Methods

Approach Best For
Standard methods (ANOVA, t-tests, AUC) Laboratory conditions, clean samples, well-characterized matrices
KWELA Environmental matrices with unknown interference, complex multi-component samples, QC assessment

Plugin Compatibility

KWELA accepts standard data frames with columns: Treatment, TTT, MS, MP, RAF. Any upstream package can produce these — use kwela_validate() to check compatibility.

# Example: quicR -> KWELA pipeline
metrics <- quicR::calculate_metrics(raw_data, meta)
kwela_validate(metrics)
result <- kwela(metrics, evaluate = TRUE)

Author

Richard A. Feiss IV, Ph.D.
Minnesota Center for Prion Research and Outreach (MNPRO)
University of Minnesota

License

MIT

About

Specialized RT-QuIC classification for complex environmental matrices

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors