PyRAD

PyRAD: RADseq Assembly and Alignment-Clustering Pipeline

PyRAD assembles de novo RADseq loci using an alignment-clustering algorithm that incorporates indel (insertion and deletion) variation and incomplete read overlap, enabling accurate identification of homologous loci across samples and diverse taxa.

Key Features:

• Alignment-Clustering Algorithm: Clusters sequences by global alignment, accommodating indel variation within and between samples and incomplete overlap among paired-end reads.
• Indel Tolerance: Integrates indel variation into locus assembly, increasing recovery of shared homologous loci across species compared to methods such as Stacks.
• Parallel Processing: Implements parallelized computation to assemble large RADseq datasets comprising hundreds of individuals.
• Hierarchical Clustering: Provides optional hierarchical clustering to organize loci into nested clusters based on similarity thresholds.
• Empirical Benchmarking: Demonstrates improved clustering of homologous loci relative to Stacks using simulated RADseq datasets containing indel variation.

Scientific Applications:

• Population Genomics: Analyzes genetic variation within and among populations using RADseq data.
• Phylogenetics: Recovers homologous loci across shallow and deep phylogenetic scales to infer evolutionary relationships.

Methodology:

Sequences are quality-filtered and clustered by global alignment under user-defined similarity thresholds. Indel variation is incorporated during clustering to prevent locus fragmentation. Parallel processing accelerates assembly, and optional hierarchical clustering refines locus grouping across multiple samples and taxa.