Hello Everyone,
in my project, I deNovo-assembled a genome of a non-model rodent using "Abyss". I now have various scaffolds, which I merged, using "GAM-NGS". I know that the genome is about 3 billion bp long. The merged genome(s) however, are much longer, some about 3.8 billion bp.
I already filtered out all scaffolds smaller than 1000 bp, which is the average length of a gene for this species. Can I naively throw out the smallest scaffolds until I reach ~3 billion bp, or do I have to tweak the merging process?
My usual way of determining the quality of an assembled genome is mapping back DNA/RNA-Seq data and look at the mapping rate, but this obviously favors larger genomes.
I would appreciate if someone could share his/her experience here.
Thanks
Robert
Update: I tested various assemblers, but Abyss performed best. I am merging assemblies for different Kmers, which is recommended in literature.
in my project, I deNovo-assembled a genome of a non-model rodent using "Abyss". I now have various scaffolds, which I merged, using "GAM-NGS". I know that the genome is about 3 billion bp long. The merged genome(s) however, are much longer, some about 3.8 billion bp.
I already filtered out all scaffolds smaller than 1000 bp, which is the average length of a gene for this species. Can I naively throw out the smallest scaffolds until I reach ~3 billion bp, or do I have to tweak the merging process?
My usual way of determining the quality of an assembled genome is mapping back DNA/RNA-Seq data and look at the mapping rate, but this obviously favors larger genomes.
I would appreciate if someone could share his/her experience here.
Thanks
Robert
Update: I tested various assemblers, but Abyss performed best. I am merging assemblies for different Kmers, which is recommended in literature.
Comment