search

marbl / merqury

k-mer based assembly evaluation
Other
290 stars 21 forks source link

evaluation for QV in haplotypes #140

Open DayTimeMouse opened 1 month ago

DayTimeMouse commented 1 month ago

Hi,

Thank you for developing this nice tool.

I have obtained hap1.fa and hap2.fa using hifiasm(HiFi +HiC reads). Next, I would like to evaluate the QV and completeness of each haplotype.

However, I am uncertain about which reads to use to build the meryl databases. Should I use haplotype-specific reads or all HiFi reads?

When I use all HiFi reads to assess the QV and completeness of each haplotype, the QV is 48, and the completeness is 90.26%. In contrast, when I use haplotype-specific reads, the QV is 36, and the completeness is 99.40%.

Best regards.

arangrhie commented 1 month ago

Hi @DayTimeMouse,

Use all HiFi reads. Completeness is measured based on num. kmers found in the assembly / found in the reads. That means, if you are using haplotype-specific reads, your denominator will be only half the genome.

Best, Arang

DayTimeMouse commented 1 month ago

Hi @Arang,

But I am still confused about using all HiFi reads to evaluate QV and completeness. Because I want to evaluate each haplotype not primary genome.

arangrhie commented 1 month ago

Run merqury with the two fasta files as input - such as merqury.sh read.meryl hap1.fa hap2.fa out. The QV file will contain QV for each haplotype and all together (both haplotypes).