Assembly Evaluation
We will now compare the results of our assembly with those of other assemblers and inspect mappings of reads to the corresponding assemblies.
MetaQUAST¶
QUAST stands for QUality ASsessment Tool. The tool evaluates genome assemblies by computing various metrics. You can find all project news and the latest version of the tool at SourceForge. QUAST is a comprehensive quality assessment tool for genome assemblies that employs several bioinformatics utilities to evaluate assembly accuracy and quality. It utilizes MUMmer (for genome alignment), GeneMarkS (self-training gene prediction in prokaryotes), GeneMark-ES (gene prediction in eukaryotes), GlimmerHMM (gene finding using Hidden Markov Models), and GAGE (a pipeline for comparing gene predictions across multiple annotations).
Additionally, MetaQUAST extends QUAST's capabilities to metagenomic assemblies by incorporating tools such as MetaGeneMark (specialized for gene prediction in metagenomes), Krona Tools (for taxonomic classification visualization), BLAST (for sequence similarity searches against reference databases), and the SILVA 16S rRNA database (for identifying microbial communities).
Task 1
Copy the pre-computed assembly results to your local directory.
cd ~/mgcourse
wget https://s3.bi.denbi.de/cmg/mgcourses/mg2025/assembly_results.tar.gz
tar -xzvf assembly_results.tar.gz
Task 2
Then copy your assembly results from the toolkit run to the assembly_results directory:
cd ~/mgcourse/
cp output/data/1/assembly/1.2.1/megahit/data_contigs.fa.gz assembly_results/megahit_out/final.contigs.fa.gz
gunzip -fd assembly_results/megahit_out/final.contigs.fa.gz
Task 3
In addition, we need to download some references in order to compare them to our assemblies:
cd ~/mgcourse/
wget https://s3.bi.denbi.de/cmg/mgcourses/mg2025/genomes.tar.gz
tar -xzvf genomes.tar.gz
To call the metaquast.py script, we have to provide reference genomes which are used to calculate a variety of different metrics for evaluating the assembly.
In real-world metagenomics, these references are not available, of course.
Task 4
Run metaquast:
cd ~/mgcourse/
metaquast.py --threads 28 --gene-finding \
-R genomes/Aquifex_aeolicus_VF5.fna,\
genomes/Bdellovibrio_bacteriovorus_HD100.fna,\
genomes/Chlamydia_psittaci_MN.fna,\
genomes/Chlamydophila_pneumoniae_CWL029.fna,\
genomes/Chlamydophila_pneumoniae_J138.fna,\
genomes/Chlamydophila_pneumoniae_LPCoLN.fna,\
genomes/Chlamydophila_pneumoniae_TW_183.fna,\
genomes/Chlamydophila_psittaci_C19_98.fna,\
genomes/Finegoldia_magna_ATCC_29328.fna,\
genomes/Fusobacterium_nucleatum_ATCC_25586.fna,\
genomes/Helicobacter_pylori_26695.fna,\
genomes/Lawsonia_intracellularis_PHE_MN1_00.fna,\
genomes/Mycobacterium_leprae_TN.fna,\
genomes/Porphyromonas_gingivalis_W83.fna,\
genomes/Wigglesworthia_glossinidia.fna \
-o quast \
-l MegaHit,metaSPAdes,Ray_31,Ray_51,velvet_31,velvet_51,idba_ud \
assembly_results/megahit_out/final.contigs.fa \
assembly_results/metaspades_out/contigs.fasta \
assembly_results/ray_31/Contigs.fasta \
assembly_results/ray_51/Contigs.fasta \
assembly_results/velvet_31/contigs.fa \
assembly_results/velvet_51/contigs.fa \
assembly_results/idba_ud_out/contig.fa
We will now inspect the QUAST reports.
Task 5
QUAST generates HTML reports including a number of interactive graphics. To access these reports, locate the html report in the quast directory and load the reports in your web browser
via the following command:
firefox quast/report.html
Task 6
Which of the assemblers performs best in terms of coverage of genome fraction of the reference genomes? Which assembly result would you prefer among them?
Solution
MEGAHIT, metaSPAdes and idba_ud perform best in terms of covered genome fraction. MEGAHIT tends to generate larger contigs along with misassemblies, while metaSPAdes creates shorter contigs with fewer misassemblies, idba_ud is quite similar to metaSPades.