MoleMap: fast alignment-free molecule mapping for long-read and linked-read sequencing data

Abstract

MotivationThe bottleneck for genome analysis will soon shift from sequencing cost to computationally expensive read alignment. When only portions of the genome are of interest, read alignment on whole-genome sequencing (WGS) data can be replaced by a fast mapping step. ResultsWe propose an approach that can map both long reads or linked-read molecules, as a pre-processing step for downstream targeted analyses. Our "molecule mapping" approach, implemented in the tool MoleMap, uses a minimized open-addressing k-mer index of the reference genome and a fast k-mer clustering procedure. We demonstrate that MoleMaps accuracy is competitive with standard alignment and mapping tools and its running time outperforms other mapping tools on 32 threads by a factor of 3 to 8 and read alignment tools by a factor of 10 to 60. Its low memory footprint allows us to analyze whole genomes on a standard laptop computer. As proofs of concept, we use MoleMap to filter reads for local assembly of a known variant region that involves non-reference sequence and showcase its use in diagnosing a patient with a rare disease. Our work contributes to more scalable genome analysis and promotes WGS for targeted analyses. Availability and ImplementationSource code of MoleMap is available at https://github.com/kehrlab/molemap.