Reconstructs metabolic pathways and estimates pathway completeness for a given set of contigs.
Go back to the main page of anvi’o programs and artifacts.
anvi-estimate-metabolism predicts the metabolic capabilities of organisms based on their genetic content. It relies upon kegg-functions and metabolism information from the KEGG resource, which is stored in a modules-db.
The metabolic pathways that this program currently considers are those defined by KOs in the KEGG MODULES resource. Each KO represents a gene function, and a KEGG module is a set of KOs that collectively carry out the steps in a metabolic pathway. Therefore, for this to work, you need to have annotated your contigs-db with hits to the KEGG KOfam database by running anvi-run-kegg-kofams prior to using this program.
Given a properly annotated contigs-db, this program determines which KOs are present and from those determines the completeness of each KEGG module. The results are described in a set of output text files, collectively referred to as kegg-metabolism.
There are several possible inputs to this program. For single genomes (isolate genomes or MAGs, for example) you can provide a contigs-db. If your contigs-db describes a metagenome rather than a single genome, you can provide the flag
--metagenome-mode. In metagenome mode, estimation is run on each contig individually - that is, only KOfam hits within the same contig are allowed to contribute to the completeness score of a given KEGG module. Alternatively, if you have binned your metagenome sequences into separate populations and would like metabolism estimation to be run separately on each bin, you can provide a profile-db and a collection.
This program always takes one or more contigs database(s) as input, but what is in those contigs dbs depends on the context (ie, genome, metagenome, bin). In the case of internal genomes (or bins), is possible to have multiple inputs but only one input contigs db. So for clarity’s sake, we sometimes refer to the inputs as ‘samples’ in the descriptions below. If you are getting confused, just try to remember that a ‘sample’ can be a genome, a metagenome, or a bin.
anvi-estimate-metabolism -c CONTIGS.db
anvi-estimate-metabolism -c CONTIGS.db --metagenome-mode
In metagenome mode, this program will estimate metabolism for each contig in the metagenome separately. This will tend to underestimate module completeness because it is likely that some modules will be broken up across multiple contigs belonging to the same population. If you prefer to instead treat all KOs in the metagenome as belonging to one collective genome, you can do so by simply leaving out the
--metagenome-mode flag (to effectively pretend that you are doing estimation for a single genome, although in your heart you will know that your contigs database really contains a metagenome). Please note that this will result in the opposite tendency to overestimate module completeness (as the KOs will in reality be coming from multiple different populations), and there will be a lot of redundancy. We are working on improving our estimation algorithm for metagenome mode. In the meantime, if you are worried about the misleading results from either of these situations, we suggest binning your metagenomes first and running estimation for the bins as described below.
You can estimate metabolism for each bin in a collection:
anvi-estimate-metabolism -c CONTIGS.db -p PROFILE.db -C COLLECTION_NAME
anvi-estimate-metabolism -c CONTIGS.db -p PROFILE.db -C COLLECTION_NAME -b BIN_NAME
Or, you can provide a specific list of bins in a text file:
anvi-estimate-metabolism -c CONTIGS.db -p PROFILE.db -C COLLECTION_NAME -B bin_ids.txt
Each line in the
bin_ids.txt file should be a bin name from the collection.
If you have a set of contigs databases of the same type (i.e., all of them are single genomes or all are binned metagenomes), you can analyze them all at once. What you need to do is put the relevant information for each contigs-db into a text file and pass that text file to anvi-estimate-metabolism. The program will then run estimation individually on each contigs database in the file. The estimation results for each database will be aggregated and printed to the same output file(s).
Multiple single genomes (also known as external-genomes) can be analyzed with the same command by providing an external genomes file to anvi-estimate-metabolism. To see the required format for the external genomes file, see external-genomes.
anvi-estimate-metabolism -e external-genomes.txt
Multiple metagenomes can be analyzed with the same command by providing a metagenomes input file. Metagenome mode will be used to analyze each contigs database in the file. To see the required format for the metagenomes file, see metagenomes.
anvi-estimate-metabolism -M metagenomes.txt
If you have multiple bins (also known as internal-genomes) across different collections or even different metagenomes, they can be analyzed with the same command by providing an internal genomes file to anvi-estimate-metabolism. To see the required format for the internal genomes file, see internal-genomes.
anvi-estimate-metabolism -i internal-genomes.txt
KEGG module completeness is computed as the percentage of steps in the metabolic pathway that are ‘present’ based on the KOs found in the contigs database. If this completeness is larger than a certain percentage, then the entire module is considered to be ‘present’ in the genome or metagenome. By default, this module completion threshold is 0.75; that is, 75 percent of the KOs in a module must have a KOfam hit in the contigs database in order for the module to be considered ‘complete’ as a whole. This threshold can be adjusted.
In this example, we change the threshold to 50 percent.
anvi-estimate-metabolism -c CONTIGS.db --module-completion-threshold 0.5
If you have previously annotated your contigs databases using a non-default KEGG data directory with
--kegg-data-dir (see anvi-run-kegg-kofams), or have moved the KEGG data directory that you wish to use to a non-default location, then you will need to specify where to find the KEGG data so that this program can use the right one. In that case, this is how you do it:
anvi-estimate-metabolism -c CONTIGS.db --kegg-data-dir /path/to/directory/KEGG
anvi-estimate-metabolism can produce a variety of output files. All will be prefixed with the same string, which by default is “kegg-metabolism”.
anvi-estimate-metabolism -c CONTIGS.db -O my-cool-prefix
Remember that module completion threshold? Well, you can use that to control which modules make it into your output files. If you provide the
--only-complete flag, then any module-related output files will only include modules that have a completeness score at or above the module completion threshold. (This doesn’t affect KO-related outputs, for obvious reasons.)
Here is an example of using this flag with long format output (which is the default, as described below, but we are asking for it explicitly here just to be clear):
anvi-estimate-metabolism -c CONTIGS.db --kegg-output-modes modules --only-complete
And here is an example of using this flag with matrix output. In this case, we are working with multiple input samples, and the behavior of this flag is slightly different: a module will be included in the matrix if it is at or above the module completion threshold in at least one sample. If there are any samples in which that module’s completeness is below the threshold, its completeness in that sample will be represented by a 0.0 in the matrix, regardless of its actual completeness score.
anvi-estimate-metabolism -i internal-genomes.txt --matrix-format --only-complete
This program has two major output options: long format (tab-delimited) output files and matrices.
Long format output has several preset “modes” as well as a “custom” mode in which the user can define the contents of the output file. Multiple modes can be used at once, and each requested “mode” will result in a separate output file. The default output mode is “modules” mode.
You can find more details on the output format by looking at kegg-metabolism.
Viewing available output modes
anvi-estimate-metabolism -c CONTIGS.db --list-available-modes
Using a non-default output mode
anvi-estimate-metabolism -c CONTIGS.db --kegg-output-modes kofam_hits
Using multiple output modes
anvi-estimate-metabolism -c CONTIGS.db --kegg-output-modes kofam_hits,modules
Viewing available output headers for ‘custom’ mode
anvi-estimate-metabolism -c CONTIGS.db --list-available-output-headers
Using custom output mode
Here is an example of defining the modules output to contain columns with the module number, the module name, and the completeness score.
anvi-estimate-metabolism -c CONTIGS.db --kegg-output-modes custom --custom-output-headers kegg_module,module_name,module_is_complete
Including modules with 0% completeness in long-format output
By default, modules with a completeness score of 0 are left out of the output files to save on space. But you can explicitly include them by adding the
anvi-estimate-metabolism -c CONTIGS.db --kegg-output-modes modules --include-zeros
Matrix format is only available when working with multiple contigs databases. Several output matrices will be generated, each of which describes one statistic such as module completion score, module presence/absence, or KO hit counts. Rows will describe modules or KOs, columns will describe your input samples (ie genomes, metagenomes, bins), and each cell of the matrix will be the corresponding statistic for a module in a sample. You can see examples of this output format by viewing kegg-metabolism.
Obtaining matrix-formatted output
anvi-estimate-metabolism -i internal-genomes.txt --matrix-format
Including KEGG metadata in the matrix output
By default, the matrix output is a matrix ready for use in other computational applications, like visualizing as a heatmap or performing clustering. But you may want to instead have a matrix that is annotated with more information, like the names and categories of each module or the functional annotations of each KO. To include this additional information in the matrix output (as columns that occur before the sample columns), use the
anvi-estimate-metabolism -i internal-genomes.txt --matrix-format --include-metadata
Note that this flag only works for matrix output because, well, the long-format output inherently includes KEGG metadata.
Note also that you can combine this flag with the
--only-complete flag, like so:
anvi-estimate-metabolism -i internal-genomes.txt --matrix-format --only-complete --include-metadata
You can see if this program is working by running the following suite of tests, which will check several common use-cases:
anvi-self-test --suite metabolism
If you have gotten an error that looks something like this:
Config Error: The contigs DB that you are working with has been annotated with a different version of the MODULES.db than you are working with now.
This means that the MODULES.db used by anvi-run-kegg-kofams has different contents (different KOs and/or different modules) than the one you are currently using to estimate metabolism, which would lead to mismatches if metabolism estimation were to continue. There are a few ways this can happen, which of course have different solutions:
--download-from-keggoptions, which get you a non-default version of KEGG data). Then you tried to run anvi-estimate-metabolism with the new kegg-data version. If this is you, and you have saved your former version of kegg-data somewhere, then you are in luck - you can simply direct anvi-estimate-metabolism to use the old version of KEGG with
--kegg-data-dir. If you didn’t save it, then unfortunately you will most likely have to re-run anvi-run-kegg-kofams on your contigs-db to re-annotate it with the new version before continuing with metabolism estimation.
--kegg-data-dirto this program.
Edit this file to update this information.
Are you aware of resources that may help users better understand the utility of this program? Please feel free to edit this file on GitHub. If you are not sure how to do that, find the
__resources__ tag in this file to see an example.