With the 36 threads of CPUs and default parameters of
ngmlr, the consuming time for mapping ~11X ONT reads to hg38 using
ngmlr0.27 (latest) is ~63 hours, while the same data to hg19 is ~ 4 hours. There is no difference in the mapping rates.
The version of hg38 is https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/001/405/GCF_000001405.39_GRCh38.p13/GRCh38_major_release_seqs_for_alignment_pipelines/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna.gz
Keeping only 24 Chromosomes and CHM, while discarding others (such as >chr1 AC:CM000663.2 gi:568336023 LN:248956422 rl:Chromosome M5:6aef897c3d6ff0c78aff06ac189178dd AS:GRCh38), would not change the mapping speed a lot.
Any ideas to speed up
nglmr mapping to hg38? Thank you.