DRAM

DRAM (Distilled and Refined Annotation of Metabolism) is a tool for annotating metagenomic assembled genomes and VirSorter identified viral contigs. DRAM annotates MAGs and viral contigs using KEGG (if provided by the user), UniRef90, PFAM, dbCAN, RefSeq viral, VOGDB and the MEROPS peptidase database as well as custom user databases. DRAM is run in two stages. First an annotation step to assign database identifiers to gene and then a distill step to curate these annotations into useful functional categories. Additionally viral contigs are further analyzed during to identify potential AMGs. This is done via assigning an auxiliary score and flags representing the confidence that a gene is both metabolic and viral.

For more detail on DRAM and how DRAM works please see our preprint as well as the wiki.

Installation

To install DRAM some dependencies need to be installed first then DRAM can be installed from this repository. In the future DRAM will be available from conda. Dependencies can be installed via conda or manually.

Conda Installation

Install DRAM into a new conda environment using the provided enviornment.yaml file.

wget https://raw.githubusercontent.com/shafferm/DRAM/master/environment.yaml
conda env create -f environment.yaml -n DRAM

If this installation method is used then all further steps should be run inside the newly created DRAM environment. This environment can be activated using this command:

conda activate DRAM

Manual Installation

If you do not install via a conda environment, then the dependencies pandas, networkx, scikit-bio, prodigal, mmseqs2, hmmer and tRNAscan-SE need to be installed manually. Then you can install DRAM using pip:

pip install DRAM-bio

Alternatively if you would like to install a development version of DRAM then you can install DRAM by cloning this repository and install using pip and the local repository.

You have now installed DRAM.

Setup

To run DRAM you need to set up the required databases in order to get annotations. All databases except for KEGG can be downloaded and set up for use with DRAM for you automatically. In order to get KEGG gene annotations and you must have access to the KEGG database. KEGG is a paid subscription service to download the protein files used by this annotator. If you do not have access to KEGG then DRAM will automatically use the KOfam HMM database to get KEGG Orthology identifiers.

I have access to KEGG

Set up DRAM using the following command:

DRAM-setup.py prepare_databases --output_dir DRAM_data --kegg_loc kegg.pep

kegg.pep is the path to the amino acid FASTA file downloaded from KEGG. This can be any of the gene fasta files that are provided by the KEGG FTP server or a concatenated version of them. DRAM_data is the path to the processed databases used by DRAM. If you already have any of the databases downloaded to your server and don't want to download them again then you can pass them to the prepare_databases command by use the --{db_name}_loc flags such as --uniref_loc and --viral_loc.

I don't have access to KEGG

Not a problem. Then use this command:

DRAM-setup.py prepare_databases --output_dir DRAM_data

Similar to above you can still provide locations of databases you have already downloaded so you don't have to do it again.

To test that your set up worked use the command DRAM-setup.py print_config and the location of all databases provided will be shown as well as the presence of additional annotation information.

NOTE: Setting up DRAM can take a long time (up to 5 hours) and uses a large about of memory (512 gb) by default. To use less memory you can use the --skip_uniref flag which will reduce memory usage to ~64 gb if you do not provide KEGG Genes and 128 gb if you do. Depending on the number of processors which you tell it to use (using the --threads argument) and the speed of your internet connection. On a less than 5 year old server with 10 processors it takes about 2 hours to process the data when databases do not need to be downloaded.

Usage

Once DRAM is set up you are ready to annotate some MAGs. The following command will generate your full annotation:

DRAM.py annotate -i 'my_bins/*.fa' -o annotation

my_bins should be replaced with the path to a directory which contains all of your bins you would like to annotated and .fa should be replaced with the file extension used for your bins (i.e. .fasta, .fna, etc). If you only need to annotated a single genome (or an entire assembly) a direct path to a nucleotide fasta should be provided. Using 20 processors DRAM.py takes about 17 hours to annotate ~80 MAGs of medium quality or higher from a mouse gut metagenome.

In the output annotation folder there will be various files. genes.faa and genes.fna are fasta files with all genes called by prodigal with additional header information gained from the annotation as nucleotide and amino acid records respectively. genes.gff is a GFF3 with the same annotation information as well as gene locations. scaffolds.fna is a collection of all scaffolds/contigs given as input to DRAM.py annotate with added bin information in the headers. annotations.tsv is the most important output of the annotation. This includes all annotation information about every gene from all MAGs. Each line is a different gene and each column contains annotation information. trnas.tsv contains a summary of the tRNAs found in each MAG.

Then after your annotation is finished you can summarize these annotations with the following command:

DRAM.py distill -i annotation/annotations.tsv -o genome_summaries --trna_path annotation/trnas.tsv --rrna_path --rrna_path annotation/rrnas.tsv

This will generate the distillate and liquor files.

System Requirements

DRAM has a large memory burden and is designed to be run on high performance computers. DRAM annotates against a large variety of databases which must be processed and stored. Setting up DRAM with KEGG Genes and UniRef90 will take up ~500 GB of storage after processing and require ~512 GB of RAM while using KOfam and skipping UniRef90 will mean all processed databases will take up ~30 GB on disk and will only use ~128 GB of RAM while processing. DRAM annotation memory usage depends on the databases used. When annotating with UniRef90 around 220 GB of RAM is required. If the KEGG gene database has been provided and UniRef90 is not used then memory usage is around 100 GB of RAM. If KOfam is used to annotate KEGG and UniRef90 is not used then less than 50 GB of RAM is required. DRAM can be run with any number of processors on a single node.

Citing DRAM

The DRAM manuscript is available as a preprint here. If DRAM helps you out in your research then please cite it.