Skip to contents

Genome annotation tools

Source: vignettes/genome_annotation.Rmd
genome_annotation.Rmd

This vignette covers the per-cell genome annotation wrappers. All of them are built on the MAP framework — see Map scripts for the underlying mechanism and how to write your own.

QUAST — assembly quality

The Bascet QUAST command computes standard per-cell assembly QC metrics from contigs, including contig counts, total length, GC content, N50/L50, NG50/LG50, and auN. It writes these metrics to the obs table of a Bascet HDF5 file.

  • Website: https://github.com/ablab/quast
  • If you use this tool, please cite: Gurevich A, Saveliev V, Vyahhi N, Tesler G. QUAST: quality assessment tool for genome assemblies. Bioinformatics. 2013;29(8):1072–1075.
  • Note - we currently only provide a partial translation of this tool for the metrics we provide. A full implementation will follow in the future

(SLURM-compatible step)

BascetMapCellQUAST(
  bascetRoot,
  inputName  = "contigs",  #or other source of contigs
  outputName = "quast",
  minContig  = 0
)

Then load the obs tables for visualization:

quast_obs <- BascetReadObsFromH5(bascetRoot, "quast")

If you already have a Seurat object, append the metrics as metadata. Cells missing from the HDF5 files receive NA; existing metadata columns with the same names are replaced.

adata <- BascetAddMetaData(
  adata,
  quast_obs
)

Abricate — AMR / virulence screening

Abricate performs mass screening of contigs against several curated databases of antimicrobial resistance and virulence genes (e.g. NCBI, CARD, ResFinder, VFDB, PlasmidFinder). It only reports acquired resistance genes, not point mutations, and is widely used because it is fast and easy to interpret.

The NCBI database is used by default. See ListDatabaseAbricate() for a list of other databases.

(SLURM-compatible step)

BascetMapCellAbricate(
  bascetRoot,
  inputName  = "contigs",  #or other source of contigs
  outputName = "abricate",
  db         = "ncbi"
)

abricate_mat <- BascetAggregateAbricate(
  bascetRoot,
  inputName = "abricate"
)

Bakta — genome annotation

Bakta is a fast, standardized annotation tool for bacterial genomes and plasmids. It identifies coding sequences, ncRNAs, tRNAs, CRISPR arrays, and more, and assigns functional descriptions through alignment-free sequence identification. It is a good choice when you want a complete genome annotation comparable across cells.

  • Website: https://github.com/oschwengers/bakta
  • If you use this tool, please cite: Schwengers O, Jelonek L, Dieckmann MA, Beyvers S, Blom J, Goesmann A. Bakta: rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microbial Genomics. 2021;7(11):000685.
  • Bakta must already be installed on the system where the job runs and available on PATH.

First download a database:

DownloadDatabaseBakta(
  dbdir  = "~/bakta",  #create directory before running command
  dbtype = "light"
)

You can then run Bakta on all cells:

(SLURM-compatible step)

BascetMapCellBakta(
  bascetRoot,
  inputName  = "contigs",  #or other source of contigs
  outputName = "bakta",
  db         = "~/bakta"
)

Then aggregate the results for visualization. This example caches the result to speed up reloading; this is optional

bakta_aggr <- BascetCacheComputation(bascetRoot,"cache_bakta",MapListAsDataFrame(BascetAggregateMap(
  bascetRoot,
  "bakta",
  aggr.bakta
)))

Ariba — AMR identification from reads

Ariba detects antimicrobial resistance genes (and other gene panels) directly from sequencing reads, without first assembling. It builds local assemblies around reference genes and reports SNPs and indels relative to the reference, which makes it useful when assembly quality is too low to trust contig-based screening.

  • Website: https://github.com/sanger-pathogens/ariba
  • If you use this tool, please cite: Hunt M, Mather AE, Sánchez-Busó L, Page AJ, Parkhill J, Keane JA, Harris SR. ARIBA: rapid antimicrobial resistance genotyping directly from sequencing reads. Microbial Genomics. 2017;3(10):e000131.
  • Ariba must already be installed on the system where the job runs and available on PATH.

(SLURM-compatible step)

BascetMapCellAriba(
  bascetRoot,
  inputName  = "filtered",
  outputName = "ariba",
  db         = "/path/to/ariba_db/out.prepareref"
)

ariba_mat <- BascetAggregateAriba(
  bascetRoot,
  inputName = "ariba"
)

GECCO — biosynthetic gene clusters

GECCO predicts biosynthetic gene clusters (BGCs) in assembled contigs using a conditional random field over Pfam domain compositions. It is much faster than antiSMASH and is well suited to scanning thousands of single-cell assemblies for natural-product potential.

  • Website: https://gecco.embl.de/
  • If you use this tool, please cite: Carroll LM, Larralde M, Fleck JS, Ponnudurai R, Milanese A, Cappio Barazzone E, Zeller G. Accurate de novo identification of biosynthetic gene clusters with GECCO. bioRxiv. 2021. doi:10.1101/2021.05.03.442509

(SLURM-compatible step)

BascetMapCellGECCO(
  bascetRoot,
  inputName  = "contigs",
  outputName = "gecco"
)

Aggregate the per-cell cluster tables into a single list of data.frames:

gecco_aggr <- BascetAggregateGECCO(
  bascetRoot,
  inputName = "gecco"
)