TMAP

TMAP aligns Ion Torrent Personal Genome Machine (PGM) short reads to reference sequences for accurate mapping and support of downstream genetic variation detection.

Key Features:

• Technology-Specific Optimization: Tuned for Ion Torrent PGM data to account for the platform's error profile and read characteristics.
• Benchmarking Framework: Incorporated into a broader procedure to evaluate mapping algorithms across high-throughput sequencing (HTS) technologies using multiple evaluation criteria.
• Computational Resource and Time Requirements: Evaluated for resource usage and runtime to balance speed and accuracy in mapping.
• Robustness of Mapping / Mapping Correctness Definition: Uses a definition of mapping correctness that accounts for expected read start and end positions and the number of insertions, deletions (indels), and substitutions.
• Repetitive Regions Handling: Reports positions for reads located in repetitive regions to improve mapping accuracy in challenging genomic contexts.
• Genetic Variation Detection: Retrieves true genetic variation positions to support variant-calling applications.
• Simulated Data Evaluation (CuReSim): Uses CuReSim to generate customized benchmark datasets by adjusting simulator parameters to specific HTS error types.
• Mapping Quality Assessment (CuReSimEval): Uses CuReSimEval to evaluate the mapping quality of reads simulated by CuReSim.
• Benchmarking of Multiple Mappers: Evaluated alongside 14 different mappers as part of its development and assessment.

Scientific Applications:

• Whole Genome Sequencing of Small Genomes: Applied to WGS projects of small genomes using Ion Torrent PGM data.
• Variant Calling and Genetic Variation Studies: Supports retrieval of true variant positions for downstream variant-calling analyses.
• Benchmarking Mapping Algorithms: Serves within benchmark procedures to compare mappers across HTS technologies.
• Mapping in Repetitive Regions: Useful for analyses requiring accurate read placement in repetitive genomic regions.

Methodology:

Development and validation used a comprehensive benchmarking process that evaluated 14 different mappers on real and simulated datasets; CuReSim generated simulated reads by adjusting parameters to specific error types, and CuReSimEval assessed the mapping quality of those simulated reads.