Readscan

Readscan identifies non-host sequences in high-throughput sequencing datasets to detect potential pathogens and estimate their relative genomic abundance.

Key Features:

• Scalability: Readscan operates as a parallel program that distributes computational tasks across multiple compute-cluster nodes to process large sequencing datasets.
• Parallel processing: The implementation analyzes multiple sequence reads simultaneously to increase throughput.
• Speed: Demonstrated on a simulated dataset of 20.1 million reads, Readscan classified human and viral sequences in under 27 minutes on a Beowulf compute cluster with 16 nodes.
• Accuracy: The tool precisely identifies non-host sequences to support reliable detection of potential pathogen-origin reads.
• Relative abundance estimation: Readscan provides estimates of the relative abundance of genomes from potential pathogens within sequencing samples.

Scientific Applications:

• Metagenomic studies: Distinguishing host and non-host sequences to characterize microbial composition in mixed samples.
• Pathogen detection and surveillance: Rapid identification of potential pathogenic sequences from high-throughput sequencing data.
• Microbial community profiling: Estimating relative genome abundances to compare community structure across samples.
• Epidemiology and environmental microbiology: Supporting analyses for infectious disease investigations and environmental sequencing surveys.

Methodology:

Parallel processing of sequencing reads by distributing computational tasks across multiple compute-cluster nodes for simultaneous analysis of reads.