Quality assessment of assemblies
Mastomys natalensis is a multimammate rodent from Sierra Leone that lives in close proximity to humans, sometimes transmitting Lassavirus to humans. In a large-scale investigation of the transmission risks of Lassavirus in Sierra Leone, different rodent species were captured/marked/released around several villages. Samples of blood (BL), from urogenital swabs (US) and oral swabs (OS) were collected and tested for Lassavirus. Lassavirus positive samples were RNA extracted and deep sequenced.
In this practical, we will assess four metagenomic assemblies (BL, OS, US and combined) from two different rodents (sample 00094 and 00187).
QUAST
Quality Assessment Tool for Genome Assemblies – QUAST/metaQUAST
Availability: http://cab.cc.spbu.ru/quast/download.html
Documentation: http://cab.cc.spbu.ru/quast/manual.html ***
1) Create a folder called validation and create links for each contig file in /home4/VBG_data/QualityAssessment/:
mkdir validation
cd validation
ln -s /home4/VBG_data/QualityAssessment/Contigs_00094/00094_* .
2) Run metaQUAST with the basic command
metaquast.py -o metaquast_00094 00094_US_L_NA_S9_contigs.fasta 00094_OS_L_NA_S2_contigs.fasta 00094_BL_L1_NA_S15_contigs.fasta 00094_all_samples_contigs.fasta
3) The output of metaQUAST will provide you with a path to the report in multiple formats. We will open report.html using firefox from the command (replace /path/to/report.html with the path printed out by metaQUAST):
firefox /path/to/report.html
Which sample provided the best assembly? Which metrics supports this? Why doesn’t the “all_samples” assembly have the largest genome?
NCBI eutilities
Useful set of commands for interfacing with NCBI directly from the command line
Installation and documentation: https://www.ncbi.nlm.nih.gov/books/NBK179288/
Helpful examples: https://bioinformatics.cvr.ac.uk/ncbi-entrez-direct-unix-e-utilities/ ***
4) It is also possible to run QUAST/metaQUAST with a reference. As we are expecting to find Lassavirus, lets download Lassavirus sequences (two segment L and S) directly from the command line using NCBI eutilites:
esearch -db nucleotide -query "OM735985" | efetch -format fasta > Lassa_L.fasta
esearch -db nucleotide -query "OM735984" | efetch -format fasta > Lassa_S.fasta
5) This time I’ll let you find the commands yourself to link to the different contig files for sample 00187, they are in /home4/VBG_data/QualityAssessment/Contigs_00094/. The syntax is the same as in 1).
6) Run the metaQUAST command, this time using as a reference Lassa_L.fasta and Lassa_S.fasta with the -R argument:
metaquast.py -l "all, US, OS, BL" 00187_all_samples_contigs.fasta 00187_US_L_NA_S26_contigs.fasta 00187_OS_L_NA_S24_contigs.fasta 00187_BL_L1_NA_S5_contigs.fasta -R Lassa_L.fasta,Lassa_S.fasta
7) Look at the report using firefox as before
Do you see a difference to the previous file? What additional metrics and visualizations are provided?
8) Try to run the metaQuast on the sample 00094 using the Lasavirus sequences as a reference with the -R.
Are there any error/warning messages during the run? Does the output look any different? What do you think is happening with this sample?
We will investigate in the metagenomic parctical what is going on with this sample.
BONUS
9) Use the different assemblies generated during the de-novo assembly practical to figure out which assembler has produced the best results? Which assembler covered the largest proportion of the genome? Which assembler produced the largest number of contigs?