Fig. 1

Workflow of overlap-based error correction using FM-index. Initially, a seed is identified from the high-quality regions with sufficient k-mer frequency. Subsequently, the flanking sequences are extended from the seed base by base using FM-index and are aligned against the query read. Note that GATA is a compression of all such strings in all reads. Finally, the most-frequent allele in each column of MSA will replace the error allele with lower frequency