GAML

GAML assembles genomes by maximizing the likelihood of candidate assemblies using a probabilistic model that accounts for sequencing error rates and insert lengths across Illumina, 454, and PacBio datasets.

Key Features:

• Probabilistic modeling: Uses a probabilistic model that captures sequencing error rates and insert lengths specific to each sequencing technology to evaluate assemblies.
• Likelihood maximization: Searches for assembly configurations that maximize the likelihood under the probabilistic model.
• Integration of diverse sequencing data: Accepts and integrates Illumina and 454 reads across various insert sizes as well as PacBio reads in a single assembly framework.
• Repeat resolution and scaffolding: Targets resolution of repeats and scaffolding of shorter contigs through its likelihood-based assembly evaluation.
• Comparative assembly performance: Achieves N50 sizes and error rates reported as comparable to established assemblers such as ALLPATHS-LG and Cerulean.

Scientific Applications:

• Multi-platform genome assembly: Integrates multiple sequencing platforms (Illumina, 454, PacBio) to produce cohesive genome assemblies from heterogeneous datasets.
• Complex genome reconstruction: Supports projects requiring repeat resolution and contig scaffolding to improve assembly continuity and accuracy.

Methodology:

Searches for an optimal assembly configuration that maximizes likelihood within a probabilistic model that explicitly accounts for dataset-specific error profiles and insert lengths.