SVM2

SVM2 detects and characterizes genomic structural variations (SVs) from ultra high-throughput genome resequencing data to improve sensitivity and specificity across diverse SV types.

Key Features:

• Integration of Multiple Analytical Methods: Combines split mapping, reassembly, read depth, and insert size analysis for comprehensive SV identification.
• Innovative Use of Insert Size Deviations: Analyzes deviations from expected library insert sizes together with local read mapping patterns to improve prediction of SV positions and types.
• Supervised Learning Approach: Employs supervised learning algorithms to refine predictions of SV position and type, increasing sensitivity compared with approaches relying solely on paired-end read mapping.
• Handling Complex Genomic Contexts: Produces reliable predictions in repetitive or low-complexity genomic regions, addressing limitations of traditional split mapping methods.

Scientific Applications:

• Genetic Diversity and Evolutionary Biology: Enables detection of SVs relevant to studies of genetic diversity and evolutionary processes.
• Disease Mechanisms and Pathogenic Conditions: Supports identification of genomic alterations that may underlie disease mechanisms or pathogenic conditions.
• Comprehensive Genome Analysis: Facilitates genome-wide SV analyses, including in repetitive or low-complexity regions where detection is challenging.

Methodology:

Integrates split mapping, reassembly, read depth, and insert size analysis; analyzes insert size deviations with local read mapping patterns; and applies supervised learning algorithms to predict SV positions and types.