search

ablab / quast

Genome assembly evaluation tool
http://quast.sf.net
Other
404 stars 78 forks source link

misassembled contig #33

Closed efratca closed 7 years ago

efratca commented 7 years ago

Hi,

I ran quast on E.coli assembly (1 contig) and I got in the contig size viewer a misassembled contig and also in the contig alignment viewer almost all the contig was misassembled. The following is the mummerplot: image Is the contig misassembled because the genome is circular? Is there a way to change the criterions for misasembly to get a correct contig?

Thanks

alexeigurevich commented 7 years ago

Hi! By default, Quast correctly processes circular genomes (does not report a misassembly), so this should not be a problem. Note: you can disable this behaviour with --eukaryote option. However, it looks like there are a lot of small discrepancies with the reference genome (it could be imperfect reference or natural diversity of the sequenced organism, e.g. structural variations). We usually call them local misassemblies. They are shown as errors (red) on Icarus contig alignment viewer by default, but you could simply hide them using top control panel: image After this, you can focus on other types of misassemblies which are more severe and may correspond to true assembly errors. As far as I can see from the mummer plot, probably all your errors are local misassemblies, so do not worry about them much. Note that you can control the threshold distinguishing local and extensive misassemblies with --extensive-mis-size option (or simply -x). The default value is 1000 bp long inconsistency.

Best, Alexey

efratca commented 7 years ago

Thanks a lot for the reply