Completeness analysis using BUSCO
Another method of assessing the quality of a genome assembly is to estimate the completeness of the genome in terms of gene sets.
BUSCO uses sets of lineage-specific single-copy core/conserved genes and searches a genome assembly for these. The output is a table of
- Complete BUSCO genes
- Complete and Single-copy
- Complete and duplicated
- Fragmented BUSCO genes
- Missing BUSCO genes
- Total BUSCO groups searched
These results give an idea of how complete the genome might be in terms of overall genes, how truncated the assembly is (# of complete/fragmented BUSCO genes), and whether there are eareas that are duplicated, e.g., due to multiple strains, wrongly assembled areas or the like.
To use BUSCO please activate the conda environment busco
course_user> conda activate busco
Change into the assembly_practical/minimap-miniasm directory and call BUSCO on the miniasm.fasta. It requires a lineage of the genome (-l flag), the assembly (–in flag), an output directory (–out flag), and we want to analyse a genome, i.e., we will specify the mode as genome (-m genome*):
course_user> busco -m genome\
course_user> -l bacteria \
coures_user> --in miniasm.fasta \
course_user> --out miniasm_busco \
course_user> -m genome
Now repeat the process with the flye assembly.
- How complete are the 2 assemblies? What are the differences?
- Overall, which assembly would you prefer?
As an interesting test you could run busco on the B. fermentans genome to get an idea what the maximum BUSCO score for this critter would be.