Hello,
I aligned (minimap2 -I 128G -ax map-hifi -t 128 -c --secondary=no assembly_hap1_hap2.fa reads.fq.gz | samtools view -bh > alignment.bam) HiFi reads to a hifiasm-generated assembly to visualize the coverage, and some contigs have very low coverage across their whole length or across multi-Mb parts:
I have about 35x coverage per haplotype, some contigs show what is expected.
The contigs that have low or no coverage have reads aligned to them, but they all have MAPQ score of 0 or 1 (MAPQ60 is 10% or less), while the ones that have 34x coverage have 1% MAPQ0 or 1, >97% MAPQ60.
There are no differences in repeat content between contigs with low or normal coverage, nor between regions at different coverage within a contig (seen with a dotplot). The behavior is not correlated with contig length, and it is consistent across all contigs within the same haplotype. Both haplotypes have the same behavior. when a contig on hap1 has low or high coverage, its homologue has the same behavior.
I run the same analysis on other assemblies, and never had this issue of many (mostly all) reads with low MAPQ score before. Do you know what is the reason for this behavior?
Hello,
I aligned (
minimap2 -I 128G -ax map-hifi -t 128 -c --secondary=no assembly_hap1_hap2.fa reads.fq.gz | samtools view -bh > alignment.bam) HiFi reads to a hifiasm-generated assembly to visualize the coverage, and some contigs have very low coverage across their whole length or across multi-Mb parts:I have about 35x coverage per haplotype, some contigs show what is expected.
The contigs that have low or no coverage have reads aligned to them, but they all have MAPQ score of 0 or 1 (MAPQ60 is 10% or less), while the ones that have 34x coverage have 1% MAPQ0 or 1, >97% MAPQ60.
There are no differences in repeat content between contigs with low or normal coverage, nor between regions at different coverage within a contig (seen with a dotplot). The behavior is not correlated with contig length, and it is consistent across all contigs within the same haplotype. Both haplotypes have the same behavior. when a contig on hap1 has low or high coverage, its homologue has the same behavior.
I run the same analysis on other assemblies, and never had this issue of many (mostly all) reads with low MAPQ score before. Do you know what is the reason for this behavior?