Exploring linkage disequilibrium in GWAS using multiple panels

LDpatterns

Explore Linkage Disequilibrium patterns given the desired settings in a large collection of LD reference panels and multiple populations

LDassoc

Interactively visualize association p-value results and linkage disequilibrium patterns for a genomic region of interest.

LDPruning

Identify independent genetic variants and prune redundant datasets based on configurable R² and p-value thresholds.

LD Visualization

Visualize LD patterns with interactive heatmaps and LD plots

Analysis Settings
1. Input Data
Load Example Data
2. Reference & Mode
3. Filters & Run


Unique Variant Count

Average R²

High LD Links


LD Result Matrix Download NxN Matrix (R²)
Pruning Configuration
1. Input Data
2. Filters & LD Parameters

Note: MAF is set to 0 for pruning.

3. Reference & Execute


Initial Variants

Variants Removed

Final Variants


Identified LD Correlations (Pairs to Check)

Final Pruned Dataset Download Pruned Data
Association Analysis Configuration
1. Data Source
Load Example Data
2. Map Columns
3. Parameters & Run

Association Data Preview
LDassoc Results

Regional Association & LD Plot

Documentation

LDSeeker is an open-source tool designed to query and compute Linkage Disequilibrium (LD) metrics across diverse human populations using multiple reference panels.

  • LDpatterns: Explore LD given a set of rsIDs (from GWAS or a list). In 'Standard' mode, it queries LD relative to the input set. In 'Pairwise' mode, it computes an NxN correlation matrix for all input variants.
  • LDassoc: It highlights an 'Index SNP' and maps all other regional SNPs relative to its LD and genomic position.
  • LDpruning: Pruning is used to identify independent genetic signals. Useful for obtaining independent sets from a GWAS.

All tools support Tab-Delimited (.txt, .tsv). Maximum file size is 500MB.

Required Columns per Module:
Module Required Fields Optional Fields
LDpatterns SNP (rsID), CHR BP (Pos)
LDassoc SNP, CHR, BP, P-value Alleles, BETA
LDpruning SNP, CHR P-value (for significance-based pruning)

Note: Genetic coordinates should be consistent with the selected Reference Panel build (e.g., hg38 for 1000G High Cov).

Linkage Disequilibrium (r²)

Measures the squared correlation of allele frequencies. Values range from 0 to 1, where 1 indicates perfect correlation.

D-Prime (D')

Normalized linkage disequilibrium. Values near 1 indicate no historical recombination between variants, even if r² is low.

Minor Allele Frequency (MAF)

Filters out variants where the minor allele occurs below a specific frequency in the population (e.g., < 0.01).

Detailed population counts and available samples for each reference panel.


Credits & Contact

LDSeeker

Main Developer

Georgios A. Manios

Computational Genetics Group

gmanios@uth.gr
Correspondence

Dr. Pantelis G. Bagos

pbagos@compgen.org

Visit GitHub Repository