Dr Yu Lin, Lecturer at the ANU Research School of Computer Science will present this seminar.
Genome assembly is a fundamental problem in computational biology and aims to reconstruct a single genome from millions or billions of short DNA segments (called reads, randomly drawn from the genome).
Metagenomics is the assembly and analysis of a mixture of multiple genomes simultaneously. It provides key insights into the composition of microbial communities found in different environments.
In this talk, we introduce novel graph models in genome assembly, which lead to substantial improvement over the state-of-the-art approaches - especially for the recent transition from short and accurate reads to long and erroneous reads. We propose to use assembly graphs to bin metagenomic sequences for the first time, which improves the typical recalls from less than 50% to above 90% and enables the detection of shared sequences among multiple genomes.
We also design new binning approaches that directly work for metagenomic long reads and help to enhance the metagenomic assembly quality, while significantly reducing the time and memory usage.