pclouds

pclouds identifies repeat structures in large eukaryotic genomes by calculating exact oligonucleotide counts to evaluate de novo repeat content.

Key Features:

• Oligonucleotide Counting Approach: Calculates exact counts for any length oligonucleotide across large genomes without relying on sequence alignment or similarity searches.
• P-cloud Construction: Builds oligonucleotide excess probability clouds ("P-clouds") that cluster related oligonucleotides occurring more frequently than expected by chance to indicate repetitive regions.
• Efficiency and Speed: Analyzes whole genomes such as the human genome on a single desktop in less than half a day, reported as at least an order-of-magnitude (~10×) speed improvement over existing methods.
• Mapping and Identification: Maps P-clouds back onto the genome and uses a sliding window approach to identify regions of high P-cloud density as repetitive regions.
• Comprehensive Detection: Detects known repeat elements and additional repetitive sequences including gene families, pseudogenes, segmental duplicons, and transposable elements (TEs), with sensitivity to small fragments (~25 bp) beyond conventional methods like RepeatMasker.
• Novel Element Identification: Develops element-specific P-clouds (ESPs) to identify novel repetitive elements such as Alu and MIR SINE sequences.

Scientific Applications:

• De Novo Repeat Structure Identification: Identifies repeat structure in newly sequenced genomes without prior repeat libraries.
• Transposable Element Analysis: Detects transposable elements and low-complexity repeats by locating clusters of high-abundance oligonucleotides.
• Genome Annotation: Predicts additional repeat-derived content, informing probabilistic genome annotation and the assessment of unannotated repetitive sequences in genomes such as human.

Methodology:

Compute exact counts for any-length oligonucleotides; construct oligonucleotide excess probability clouds (P-clouds); map P-clouds onto the genome and apply a sliding window to identify high-density repetitive regions; generate element-specific P-clouds (ESPs) for novel Alu and MIR SINE identification.