Changelog

All notable changes to edge-gwas will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

Version 0.1.1 (2025-12-25) - Current Release

Status: ⚠️ Public Testing Phase

Added

Population Structure Control:

  • load_grm_gcta() - Load GRM calculated by GCTA

  • calculate_grm_gcta() - Calculate GRM using GCTA

  • GRM integration in calculate_alpha() and apply_alpha() for mixed model analysis

  • identify_related_samples() - Identify related sample pairs from GRM

  • filter_related_samples() - Remove related samples

Principal Component Analysis:

  • calculate_pca_plink() - PCA using PLINK2 with LD pruning

  • calculate_pca_pcair() - PC-AiR for related samples

  • calculate_pca_sklearn() - Basic PCA using scikit-learn

  • attach_pcs_to_phenotype() - Merge PCs with phenotype data

  • get_pc_covariate_list() - Helper for generating PC covariate names

Outcome Transformations:

  • outcome_transform parameter in EDGEAnalysis for continuous outcomes

  • Support for ‘log’, ‘log10’, ‘inverse_normal’, and ‘rank_inverse_normal’ transformations

Additional Data Formats:

  • load_pgen_data() - Load PLINK2 .pgen/.pvar/.psam files

  • load_bgen_data() - Load BGEN format with dosages

  • load_vcf_data() - Load VCF/VCF.gz files

Tool Installation:

  • edge-gwas-install-tools - Interactive installer for PLINK2, GCTA, and R packages

  • edge-gwas-check-tools - Verify external tool installations

  • Support for Linux and macOS (including ARM64)

Quality Control:

  • filter_samples_by_call_rate() - Filter samples by genotype call rate

  • calculate_hwe_pvalues() - Calculate Hardy-Weinberg Equilibrium p-values

  • filter_variants_by_hwe() - Filter variants by HWE p-value

  • check_case_control_balance() - Check case/control ratio

Analysis Methods:

  • cross_validated_edge_analysis() - K-fold cross-validation for EDGE

  • additive_gwas() - Standard additive model for comparison

Changed

Data Handling:

  • Breaking Change: Replaced Koalas DataFrames with pandas DataFrames throughout codebase

  • Updated load_participant_data() to use pandas instead of participant.retrieve_fields().to_koalas()

  • All PCA functions now return DataFrames with ‘IID’ column and IID as index

  • Consistent sample ID handling (all IDs converted to strings for matching)

Core Analysis:

  • EDGEAnalysis class now supports GRM as optional input in all methods

  • calculate_alpha() and apply_alpha() accept grm_matrix and grm_sample_ids parameters

  • Linear mixed model implementation for continuous outcomes with GRM

  • Logistic mixed model for binary outcomes with GRM

Dependencies:

  • Removed dependency on databricks.koalas

  • All operations now use native pandas (>= 1.2.0)

  • Added optional dependencies: pgenlib, bgen-reader, cyvcf2

  • External tool requirements: PLINK2, GCTA, R (all optional)

Migration from v0.1.0

Breaking Changes:

  • Code using .to_koalas() must be updated to use pandas DataFrames

  • Remove any import databricks.koalas statements

  • PCA functions now return DataFrames indexed by IID

Migration Example:

# Old code (v0.1.0)
import databricks.koalas as ks
participant_df = participant.retrieve_fields(fields=fields, engine=dxdata.connect()).to_koalas()

# New code (v0.1.1)
import pandas as pd
participant_df = participant.retrieve_fields(fields=fields, engine=dxdata.connect())

# New features in v0.1.1
from edge_gwas.utils import calculate_grm_gcta, load_grm_gcta, calculate_pca_plink

# Calculate GRM for population structure control
grm_prefix = calculate_grm_gcta('genotypes', maf_threshold=0.01)
grm_matrix, grm_ids = load_grm_gcta(grm_prefix)

# Calculate PCA
pca_df = calculate_pca_plink('genotypes', n_pcs=10)

# Use outcome transformation and GRM in analysis
edge = EDGEAnalysis(outcome_type='continuous', outcome_transform='rank_inverse_normal')
alpha_df, gwas_df = edge.run_full_analysis(
    train_g, train_p, test_g, test_p,
    outcome='trait', covariates=['age', 'sex', 'PC1', 'PC2', 'PC3'],
    grm_matrix=grm_matrix, grm_sample_ids=grm_ids
)

Bug Fixes

  • Resolved compatibility issues with recent pandas versions

  • Improved memory efficiency in data loading operations

  • Fixed GRM sample alignment when samples differ between genotype and GRM

  • Fixed ID type inconsistencies (all IDs now converted to strings)

  • Improved convergence in mixed model fitting

Version 0.1.0 (2025-12-24)

Status: Superseded by v0.1.1

This is the first packaged release of EDGE-GWAS, transitioning from standalone scripts (v0.0.0) to a proper Python package.

Added

Core Functionality:

  • EDGEAnalysis class for two-stage GWAS analysis

  • calculate_alpha() - Calculate encoding parameters from training data

  • apply_alpha() - Apply encoding to test data for GWAS

  • run_full_analysis() - Complete end-to-end workflow

  • Support for binary outcomes (logistic regression)

  • Support for quantitative outcomes (linear regression)

Data Handling:

  • load_plink_data() - Load PLINK binary format (.bed/.bim/.fam)

  • prepare_phenotype_data() - Load and prepare phenotype files

  • stratified_train_test_split() - Train/test data splitting with stratification

  • filter_variants_by_maf() - Minor allele frequency filtering

  • filter_variants_by_missing() - Missingness filtering

Statistical Functions:

  • calculate_genomic_inflation() - Calculate genomic inflation factor (λ)

  • Two-stage analysis framework (calculate alpha → apply alpha)

  • Parallel processing support with configurable CPU cores

  • Convergence monitoring and skipped SNP tracking

Visualization:

  • manhattan_plot() - Create Manhattan plots with customization

  • qq_plot() - Create QQ plots with lambda calculation

  • plot_alpha_distribution() - Visualize alpha value distribution

Documentation:

  • Complete Sphinx documentation with Read the Docs integration

  • Installation guide, quick start guide, and API reference

Dependencies

  • numpy >= 1.19.0

  • pandas >= 1.2.0

  • scipy >= 1.6.0

  • statsmodels >= 0.12.0

  • scikit-learn >= 0.24.0

  • matplotlib >= 3.3.0

  • pandas-plink >= 2.0.0

  • databricks-koalas (deprecated in v0.1.1)

—

Version 0.0.0 (2024-04-02) - Legacy

Status: 🔒 Deprecated - No Maintenance

Original EDGE implementation as standalone Python scripts.

Warning

Version 0.0.0 is deprecated and no longer maintained. Users should migrate to v0.1.1 or later.

—

Version History Summary

Version

Date

Status

Key Features

0.1.1

2025-12-25

Current

GRM support, PCA methods, outcome transformations, pandas migration

0.1.0

2025-12-24

Superseded

First packaged release, core EDGE functionality

0.0.0

2024-04-02

Deprecated

Original standalone scripts

—

See Also

Documentation:

—

Last updated: 2025-12-25 for edge-gwas v0.1.1

For questions or issues, visit: https://github.com/nicenzhou/edge-gwas/issues