FiberGrowth

Short description

FiberGrowth allows to detect specific Polysaccharides Utilization Loci (PUL) from strain's genome.

Usage

Rscript FiberGrowth.R [options] --out=<output_folder> --gff=<gff_file> --proteins=<fasta_file>

Rscript FiberGrowth.R [options] --out=<output_folder> --genome=<fasta_file>

Rscript FiberGrowth.R (-h | --help)

Options:
  --gff=<gff_file>                          Gene locations in gff format.
  --proteins=<fasta_file>                   Amino acid sequences in fasta format.
  --genome=<fasta_file>                     Genome in fasta format.
  --lib=<path_to_library>                   Path to library of PUL models
                                            [default: PUL_models in installation directory].
  --out=<output_folder>                     Output directory (will be created).
  --gffFeatureName=<gffFeatureName>         Name of feature to use in gff file [default: CDS]
  --gffAttributeKey=<gffAttributeKey>       Name of attribute key to use in gff file [default: Name]
  -t --threads=<threads>                    Number of threads to run [default: 1].
  -h, --help                                Show help.

Installation

Dependencies

• Hmmer (>3.1b2) (http://hmmer.org)
• Prodigal (>2.6.3) (https://github.com/hyattpd/Prodigal)
• Pandoc (>2.14)
• R (>4)
• R packages:
  • data.table
  • DT
  • docopt
  • gggenes
  • ggplot2
  • knitr
  • magrittr
  • rhmmer
  • rmarkdown

Quick start

Running FiberGrowth on a given genome sequence in fasta format:

Rscript FiberGrowth.R --out=fibergrowth_output --genome=genome.fasta

Alternatively, FiberGrowth can be run on amino acid sequences in fasta format and gene locations in gff format:

Rscript FiberGrowth.R --out=fibergrowth_output --gff=genome.gff --proteins=proteins.faa

Example: Predicting PULs on B. thetaiotaomicron VPI-5482

Download the assembly of Bacteroides thetaiotaomicron VPI-5482 from NCBI:

curl https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/011/065/GCF_000011065.1_ASM1106v1/GCF_000011065.1_ASM1106v1_genomic.fna.gz | gunzip > GCF_000011065.1_ASM1106v1_genomic.fna

Run FiberGrowth:

Rscript FiberGrowth.R --out=example --genome=GCF_000011065.1_ASM1106v1_genomic.fna

Inspect the output: For each fiber a table with candidate PULs is created:

# list all tables:
ls -l example/fibergrowth
# inspect results for heparin:
less example/fibergrowth/gn_heparin_fibergrowth.tsv

In addition, a report in html format is generated:

example/FiberGrowthReport.html

Custom PUL models

It is possible to run FiberGrowth with custom PUL models in addition to the models in PUL_models. The following files are required:

• protein_families.hmm: HMMs representing the protein families of the PUL. HMMs can be custom build using hmmbuild (HMMER3 package) or retrieved from PFAM. Each HMM requires the following tags in the header section:

  • NAME: A unique name of the protein family model.
  • GA: Gathering thresholds. Hits with a score below GA will be ignored.

• .h3i, .h3f, .h3p, .h3m: In order to be used in FiberGrowth, hmmpress (HMMER3 package) has to be applied on the protein_families.hmm file.

• pathway_table.tsv: A tab separated file containing the metadata of the PUL model with the columns:

  • HmmID: Name of the protein family. Has to match the NAME tag in the respective HMM.
  • PathwayName: Name of the pathway/PUL.
  • Position: Position of the gene in the PUL.
  • RefStrand: Orientation of the gene (not used in the current implementation).
  • GeneName: Name of the gene. Will be included in output tables and plots.
  • Weight: Indicates core genes. A PUL will only be reported if all genes with Weight == 1 are present. Set Weight to 0 for all genes to report all PULs above MinSize independently of composition.
  • MinSize: Minimum gene number required for a PUL to be reported.

All files have to be stored in a directory with a name matching the PathwayName. For each PUL model, a separate directory is required. See PUL_models for examples. For more details on HMMs, please see the HMMER3 documentation.

References

Besides custom built protein HMMs, FiberGrowth uses HMMs retrieved from Pfam (http://pfam.xfam.org; El-Gebali et al., 2019). The HMMs are part of this repository and can be found in the PUL_models subdirectory.