Figbird

Figbird fills gaps in draft genome assemblies using a probabilistic generative model applied to second-generation sequencing reads to improve assembly completeness and accuracy.

Key Features:

• Probabilistic gap filling: Uses a generative model for sequencing that incorporates insert size distributions and sequencing error models to predict and fill assembly gaps.
• Expectation-Maximization algorithm: Implements an expectation-maximization (EM) algorithm that iteratively refines estimates for missing sequence and allocates reads across gaps.
• Second-generation sequencing support: Operates on second-generation sequencing reads as the input data for gap inference and filling.
• Error minimization: Fills large portions of gaps while producing a minimal number of errors and misassemblies in the filled sequences.
• Implementation: Implemented in C++ for computational performance.

Scientific Applications:

• Genome assembly improvement: Enhances the completeness and continuity of draft genomes, particularly for assemblies with gaps.
• Complex-genome projects: Applies to genomes with complex structures or limited sequencing resources where gaps are prevalent.
• Comparative genomics: Produces more complete assemblies for comparative genomic analyses.
• Evolutionary studies: Supports evolutionary investigations by reducing missing sequence in draft assemblies.
• Functional genomics: Improves utility of draft genomes for functional genomic analyses by increasing contiguous sequence.

Methodology:

Employs a sequencing generative model that incorporates insert size distributions and sequencing error models and applies an expectation-maximization (EM) algorithm to iteratively refine missing-sequence estimates and distribute reads across gaps; implemented in C++.