Allele-Specific Binding (ASB)

ASB analysis leverages heterozygosity as a built-in experimental control. Because the maternal and paternal alleles reside in the exact same cellular environment and are exposed to identical concentrations of transcription factors, any observed difference in binding affinity can be attributed directly to the sequence variant itself. This eliminates confounding variables like cell cycle state, trans-factor concentration, and batch effects that plague traditional ChIP-seq comparisons between individuals.

The core computational task is counting sequencing reads that overlap heterozygous single nucleotide variants and testing for allelic imbalance. Key steps include:

• WASP mapping: Corrects for reference allele mapping bias by swapping alleles and remapping discarded reads
• Binomial exact test: Tests the null hypothesis of 50/50 allelic distribution against observed counts
• Benjamini-Hochberg correction: Controls false discovery rate across thousands of tested heterozygous sites
• Minimum read depth filtering: Requires sufficient coverage (typically ≥20 reads) for statistical power

ASB provides the missing mechanistic link between GWAS loci and disease biology. A non-coding variant identified through genome-wide association studies may reside in an enhancer, but ASB analysis demonstrates whether that variant actually disrupts transcription factor binding. This transforms statistical associations into causal regulatory hypotheses. For example, an ASB event at a type 2 diabetes GWAS locus in pancreatic islet cells can pinpoint the exact variant that alters FOXA2 binding affinity.

Modern ASB analysis integrates sequence-based deep learning models like BPNet and Enformer to predict allelic binding differences in silico. These models compute the predicted binding signal for both the reference and alternate allele sequences, generating an allelic effect score. Strong concordance between predicted and observed allelic imbalance validates the model's ability to learn transcription factor binding grammar and enables high-throughput variant prioritization across entire genomes without requiring individual ChIP-seq experiments.

ASB events typically arise from variants that directly alter the core binding motif of a transcription factor. A single nucleotide change can:

• Abolish binding: Disrupt a critical base contact in the major groove
• Create a novel motif: Generate a binding site for a different factor
• Modulate affinity: Subtly shift the binding energy landscape without complete loss Position weight matrix (PWM) scanning of reference and alternate alleles quantifies the change in motif match score, providing a mechanistic explanation for the observed allelic imbalance.

ASB is not a static property of a variant but is highly context-dependent. The same heterozygous variant may exhibit strong allelic imbalance in one cell type but equal binding in another, reflecting tissue-specific transcription factor expression. This property makes ASB a powerful tool for identifying the causal cell type for disease-associated variants. A variant showing ASB exclusively in microglia but not neurons implicates microglial regulatory programs in neurodegenerative disease pathogenesis.