Should N50 change after JuiceBox edditing

[kkrizanovic]

Hi,

I'm scaffolding a quite large plant genome (estimated 4.1 Gbp). I've used hifiasm to get an initial assembly from PacBio HiFi reads, and then 3d-dna pipeline to get initial scaffolds. Initial scaffolds have N50 of about 100Mbp.

After editing the scaffolds using Juicebox (the heatmap and the assembly file), I've created a new FASTA file using run-asm-pipeline-post-review.sh script as per instructions in Genome assembly cookbook.

After manual editing with Juicebox, the N50 measure increased to about 200Mbp. Is that OK? I would expect the N50 to remain the same. I did not merge any contigs in Juicebox only rearanged their positions and maybe split few of them. I've changed chromosome definitions to match 13 chromosomes that were evident from the edited heatmap.

For N50 calculations I've used Quast.

Krešimir Križanović