Comparison of BWA-MEM and NovoAlign using human whole-genome next-generation sequencing reads

The development of sequencing technology has led to an increasing amount of next-generation sequencing data．To align numerous reads to reference genomes accurately is the basis for downstream analysis．BWA-MEM and NovoAlign, the most widely used DNA-seq alignment tools, have not been evaluated for applications to different structural regions in a genome．In the present work, we estimate the two alignment tools in low complexity region, segmental duplication region and the remaining region of the human genome using real and simulated data．BWA-MEM could align reads to reference genome, and even excessively align reads to low complexity regions and segmental duplication regions with low mapping quality．Compared with BWA-MEM, NovoAlign could align relatively fewer reads to reference genome, but most aligned reads have higher mapping quality．It is suggested that NovoAlign should be used when genomes are interspersed high segmental duplications．