Mako

Mako detects and characterizes complex structural variants (CSVs) in genomic sequences to enable analysis of intricate breakpoint connections and investigation of CSV formation mechanisms.

Key Features:

• CSV detection: Identifies complex structural variants defined by more than two breakpoints in genomic sequences.
• Bottom-up model-free strategy: Employs a bottom-up guided model-free approach rather than relying on predefined models.
• Graph-based pattern growth: Uses a graph-based pattern growth algorithm to discover variant patterns.
• Mutational signal graph representation: Constructs a mutational signal graph where potential breakpoint connections are represented as nodes and edges.
• Maximal subgraph identification: Detects maximal subgraphs in the mutational signal graph that correspond to CSV events.
• Breakpoint characterization: Accurately characterizes intricate breakpoint connections within CSV events.
• CSV type discovery: Identified 15 distinct types of CSVs, including two novel forms: adjacent segment swap and tandem dispersed duplication.
• Performance on datasets: Demonstrated consistent outperformance of existing algorithms on both simulated and real datasets.
• Validation metrics: Achieved validation rates around 70% based on experimental and computational validations and manual inspections.
• Breakpoint shift metrics: Reported median breakpoint shifts of 13 base pairs (bp) experimentally and 26 bp computationally.
• Sequence homology analysis: Provides analyses that shed light on the role of sequence homology in CSV formation.

Scientific Applications:

• Structural variation discovery: Detection and detailed characterization of complex structural variants in genomic data.
• Mechanistic studies: Investigation of CSV formation mechanisms, including the role of sequence homology.
• Benchmarking and evaluation: Comparative evaluation of CSV detection performance using simulated and real datasets.
• Genetic diversity and disease research: Study of structural variation impacts on genetic diversity and disease.

Methodology:

Constructs a mutational signal graph of potential breakpoint connections and applies a bottom-up guided, graph-based pattern growth procedure to identify maximal subgraphs corresponding to CSVs; evaluated on simulated and real datasets.