Abstract
Metagenomic sequence classification should be fast, accurate and information-rich. Emerging long-read sequencing technologies promise to improve the balance between these factors but most existing methods were designed for short reads. MetaMaps is a new method, specifically developed for long reads, that combines the accuracy of slower alignment-based methods with the scalability of faster k-mer-based methods. Using an approximate mapping algorithm, it is capable of mapping a long-read metagenome to a comprehensive RefSeq database with >12,000 genomes in <30 GB or RAM on a laptop computer. Integrating these mappings with a probabilistic scoring scheme and EM-based estimation of sample composition, MetaMaps achieves >95% accuracy for species-level read assignment and r2 > 0.98 for the estimation of sample composition on both simulated and real data. Uniquely, MetaMaps outputs mapping locations and qualities for all classified reads, enabling functional studies (e.g. gene presence/absence) and the detection of novel species not present in the current database.
Availability and Implementation MetaMaps is implemented in C++/Perl and freely available from https://github.com/DiltheyLab/MetaMaps (GPL v3).