nf-core/taxprofiler is a pipeline for highly-parallelised taxonomic classification and profiling of shotgun metagenomic data across multiple tools simultaneously. In addition to multiple classification and profiling tools, at the same time it allows you to performing taxonomic classification and profiling across multiple databases and settings per tool, as well as produces standardised output tables to allow immediate cross comparison of results between tools.
When running nf-core/taxprofiler, every step and tool is 'opt in'. To run a given classifier or profiler you must make sure to supply both a database in your `<database>.csv` and supply `--run_<profiler>` flag to your command. Omitting either will result in the profiling tool not executing.
nf-core/profiler also includes optional pre-processing (adapter clipping, merge running etc.) or post-processing (visualisation) steps. These are also opt in with a `--perform_<step>` flag. In some cases, the pre- and post-processing steps may also require additional files. Please check the parameters tab of this documentation for more information.
Please see the rest of this page for information about how to prepare input samplesheets and databases and how to run Nextflow pipelines. See the [parameters](https://nf-co.re/taxprofiler/parameters) documentation for more information about specific options the pipeline also offers.
nf-core/taxprofiler can accept as input raw or preprocessed single- or paired-end short-read (e.g. Illumina) FASTQ files, long-read FASTQ files (e.g. Oxford Nanopore), or FASTA sequences (available for a subset of profilers).
You will need to create a samplesheet with information about the samples you would like to analyse before running the pipeline. Use this parameter to specify its location. It has to be a comma-separated file with 6 columns, and a header row as shown in the examples below. Furthermother, nf-core/taxprofiler also requires a second comma-separated file of 3 columns with a header row as in the examples below.
This samplesheet is then specified on the command line as follows:
The `sample` identifiers have to be the same when you have re-sequenced the same sample more than once e.g. to increase sequencing depth. The pipeline will concatenate different runs FASTQ files of the same sample before performing profiling, when `--perform_runmerging` is supplied. Below is an example for the same sample sequenced across 3 lanes:
> ⚠️ Runs of the same sample sequenced on Illumina platforms with a combination of single and paired-end data will **not** be run-wise concatenated, unless pair-merging is specified. In the example above, `run3` will be profiled independently of `run1` and `run2` if pairs are not merged.
The pipeline will auto-detect whether a sample is single- or paired-end using the information provided in the samplesheet. The samplesheet can have as many columns as you desire, however, there is a strict requirement for the first 6 columns to match those defined in the table below.
A final samplesheet file consisting of both single- and paired-end data, as well as long-read FASTA files may look something like the one below. This is for 6 samples, where `2612` has been sequenced twice.
> ⚠️ While one can include both short-read and long-read data in one run, we recommend that you split these across _two_ pipeline runs and database sheets (see below). This will allow classification optimisation for each data type, and make MultiQC run-reports more readable (due to run statistics having vary large number differences).
| `run_accession` | Run ID or name unique for each (pairs of) file(s) .Can also supply sample name again here, if only a single run was generated [required]. |
| `instrument_platform` | Sequencing platform reads generated on, selected from the EBI ENA [controlled vocabulary](https://www.ebi.ac.uk/ena/portal/api/controlledVocab?field=instrument_platform) [required]. |
| `fastq_1` | Path or URL to sequencing reads or for Illumina R1 sequencing reads in FASTQ format. GZipped compressed files accepted. Can be left empty if data in FASTA is specifed. Cannot be combined with `fasta`. |
| `fastq_2` | Path or URL to Illumina R2 sequencing reads in FASTQ format. GZipped compressed files accepted. Can be left empty if single end data. Cannot be combined with `fasta`. |
| `fasta` | Path or URL to long-reads or contigs in FASTA format. GZipped compressed files accepted. Can be left empty if data in FASTA is specifed. Cannot be combined with `fastq_1` or `fastq_2`. |
Databases can be supplied either in the form of a compressed `.tar.gz` archive of a directory containing all relevant database files or the path to a directory on the filesystem.
> ⚠️ nf-core/taxprofiler does not provide any databases by default, nor does it currently generate them for you. This must be performed manually by the user. See below for more information of the expected database files.
The pipeline takes the paths and specific classification/profiling parameters of the tool of these databases as input via a four column comma-separated sheet.
> ⚠️ To allow user freedom, nf-core/taxprofiler does not check for mandatory or the validity of non-file database parameters for correct execution of the tool - excluding options offered via pipeline level parameters! Please validate your database parameters (cross-referencing [parameters](https://nf-co.re/taxprofiler/parameters, and the given tool documentation) before submitting the database sheet! For example, if you don't use the default read length - Bracken will require `-r <read_length>` in the `db_params` column.
An example database sheet can look as follows, where 5 tools are being used, and `malt` and `kraken2` will be used against two databases each. This is because specifying `bracken` implies first running `kraken2` on the same database.
For Bracken, if you wish to supply any parameters to either the Kraken or Bracken step you **must** have a _semi-colon_`;` list as in `db_params`. This is to allow to specify the Kraken2 parameters before, and Bracken parameters after the `;` as Bracken is a two step process. This is particularly important if you supply a Bracken database with a non-default read length parameter. If you do not have any parameters to specify, you can leave this as empty.
| `tool` | Taxonomic profiling tool (supported by nf-core/taxprofiler) that the database has been indexed for [required]. Please note that `bracken` also implies running `kraken2` on the same database. |
| `db_name` | A unique name per tool for the particular database [required]. Please note that names need to be unique across both `kraken2` and `bracken` as well, even if re-using the same database. |
| `db_params` | Any parameters of the given taxonomic classifier/profiler that you wish to specify that the taxonomic classifier/profiling tool should use when profiling against this specific database. Can be empty to use taxonomic classifier/profiler defaults. Must not be surrounded by quotes [required]. We generally do not recommend specifying parameters here that turn on/off saving of output files or specifying particular file extensions - this should be already addressed via pipeline parameters. For Bracken databases, must at a minimum contain a `;` separating Kraken2 from Bracken parameters. |
| `db_path` | Path to the database. Can either be a path to a directory containing the database index files or a `.tar.gz` file which contains the compressed database directory with the same name as the tar archive, minus `.tar.gz` [required]. |
> 💡 You can also specify the same database directory/file twice (ensuring unique `db_name`s) and specify different parameters for each database to compare the effect of different parameters during classification/profiling.
- [**Bracken**:](#bracken-custom-database) output of the combined `kraken2-build` and `bracken-build` process.
- [**Centrifuge**:](#centrifuge-custom-database) output of `centrifuge-build`.
- [**DIAMOND**:](#diamond-custom-database) output of `diamond makedb`.
- [**Kaiju**:](#kaiju-custom-database) output of `kaiju-makedb`.
- [**Kraken2**:](#kraken2-custom-database) output of `kraken2-build` command(s).
- [**KrakenUniq**:](#krakenuniq-custom-database) output of `krakenuniq-build` command(s).
- [**MALT**](#malt-custom-database) output of `malt-build`.
- [**MetaPhlAn3**:](#metaphlan3-custom-database) output of with `metaphlan --install` or downloaded from links on the [MetaPhlAn3 wiki](https://github.com/biobakery/MetaPhlAn/wiki/MetaPhlAn-3.0#customizing-the-database).
- [**mOTUs**:](#motus-custom-database) is composed of code and database together.
When running nf-core/taxprofiler, every step and tool is 'opt in'. To run a given classifier/profiler you must make sure to supply both a database in your `<database>.csv` and supply `--run_<profiler>` flag to your command. Omitting either will result in the classification/profiling tool not executing. If you wish to perform pre-processing (adapter clipping, merge running etc.) or post-processing (visualisation) steps, these are also opt in with a `--perform_<step>` flag. In some cases, the pre- and post-processing steps may also require additional files. Please check the parameters tab of this documentation for more information.
[`FastQC`](https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) gives general quality metrics about your reads. It provides information about the quality score distribution across your reads, per base sequence content (%A/T/G/C), adapter contamination and overrepresented sequences. nf-core taxprofiler offers [`falco`](https://github.com/smithlabcode/falco) as an drop-in replacement, with supposedly better improvement particularly for long reads.
Raw sequencing read processing in the form of adapter clipping and paired-end read merging can be activated via the `--perform_shortread_qc` or `--perform_longread_qc` flags.
It is highly recommended to run this on raw reads to remove artifacts from sequencing that can cause false positive identification of taxa (e.g. contaminated reference genomes) and/or skews in taxonomic abundance profiles. If you have public data, normally these should have been corrected for, however you should still check that these steps have indeed been already performed.
There are currently two options for short-read preprocessing: [`fastp`](https://github.com/OpenGene/fastp) or [`adapterremoval`](https://github.com/MikkelSchubert/adapterremoval).
For adapter clipping, you can either rely on the tool's default adapter sequences, or supply your own adapters (`--shortread_qc_adapter1` and `--shortread_qc_adapter2`)
By default, paired-end merging is not activated. In this case paired-end 'alignment' against the reference databases is performed where supported, and if not, supported pairs will be independently classified/profiled. If paired-end merging is activated you can also specify whether to include unmerged reads in the reads sent for classification/profiling (`--shortread_qc_mergepairs` and `--shortread_qc_includeunmerged`).
You can also turn off clipping and only perform paired-end merging, if requested. This can be useful when processing data downloaded from the ENA, SRA, or DDBJ (`--shortread_qc_skipadaptertrim`).
Both tools support length filtering of reads and can be tuned with `--shortread_qc_minlength`. Performing length filtering can be useful to remove short (often low sequencing complexity) sequences that result in unspecific classification and therefore slow down runtime during classification/profiling, with minimal gain.
Complexity filtering is primarily a run-time optimisation step. It is not necessary for accurate taxonomic classification/profiling, however it can speed up run-time of each tool by removing reads with low-diversity of nucleotides (e.g. with mono-nucleotide - `AAAAAAAA`, or di-nucleotide repeats `GAGAGAGAGAGAGAG`) that have a low-chance of giving an informative taxonomic ID as they can be associated with many different taxa. Removing these reads therefore saves computational time and resources.
There are currently three options for short-read complexity filtering: [`bbduk`](https://jgi.doe.gov/data-and-tools/software-tools/bbtools/bb-tools-user-guide/bbduk-guide/), [`prinseq++`](https://github.com/Adrian-Cantu/PRINSEQ-plus-plus), and [`fastp`](https://github.com/OpenGene/fastp#low-complexity-filter).
The tools offer different algorithms and parameters for removing low complexity reads and quality filtering. We therefore recommend reviewing the pipeline's [parameter documentation](https://nf-co.re/taxprofiler/parameters) and the documentation of the tools (see links above) to decide on optimal methods and parameters for your dataset.
You can optionally save the FASTQ output of the run merging with the `--save_complexityfiltered_reads`. If running with `fastp`, complexity filtering happens inclusively within the earlier shortread preprocessing step. Therefore there will not be an independent pipeline step for complexity filtering, and no independent FASTQ file (i.e. `--save_complexityfiltered_reads` will be ignored) - your complexity filtered reads will also be in the `fastp/` folder in the same file(s) as the preprocessed read.
> ⚠️ For nanopore data: we do not recommend performing any read preprocessing or complexity filtering if you are using ONTs Guppy toolkit for basecalling and post-processing.
Removal of possible-host reads from FASTQ files prior classification/profiling can be activated with `--perform_shortread_hostremoval` or `--perform_longread_hostremoval`.
Similarly to complexity filtering, host-removal can be useful for runtime optimisation and reduction in misclassified reads. It is not always necessary to report classification of reads from a host when you already know the host of the sample, therefore you can gain a run-time and computational advantage by removing these prior typically resource-heavy classification/profiling with more efficient methods. Furthermore, particularly with human samples, you can reduce the number of false positives during classification/profiling that occur due to host-sequence contamination in reference genomes on public databases.
nf-core/taxprofiler currently offers host-removal via alignment against a reference genome with Bowtie2 for short reads and minimap2 for long reads, and the use of the unaligned reads for downstream classification/profiling.
You can supply your reference genome in FASTA format with `--hostremoval_reference`. You can also optionally supply a directory containing pre-indexed Bowtie2 index files with `--shortread_hostremoval_index` or a minimap2 `.mmi` file for `--longread_hostremoval_index`, however nf-core/taxprofiler will generate these for you if necessary. Pre-supplying the index directory or files can greatly speed up the process, and these can be re-used.
> 💡 If you have multiple taxa or sequences you wish to remove (e.g., the host genome and then also PhiX - common quality-control reagent during sequencing) you can simply concatenate the FASTAs of each taxa or sequences into a single reference file.
#### Run Merging
For samples that may have been sequenced over multiple runs, or for FASTQ files split into multiple chunks, you can activate the ability to merge across all runs or chunks with `--perform_runmerging`.
For more information how to set up your input samplesheet, see [Multiple runs of the same sample](#multiple-runs-of-the-same-sample).
Activating this functionality will concatenate the FASTQ files with the same sample name _after_ the optional preprocessing steps and _before_ classification/profiling. Note that libraries with runs of different pairing types will **not** be merged and this will be indicated on output files with a `_se` or `_pe` suffix to the sample name accordingly.
The following sections provide tips and suggestions for running the different taxonomic classification and profiling tools _within the pipeline_. For advice and/or guidance whether you should run a particular tool on your specific data, please see the documentation of each tool!
An important distinction between the different tools in included in the pipeline is classification versus profiling. Taxonomic _classification_ is concerned with simply detecting the presence of species in a given sample. Taxonomic _profiling_ involves additionally estimating the _abundance_ of each species.
Note that not all taxonomic classification tools (e.g. Kraken, MALT, Kaiju) performs _profiling_, but all taxonomic profilers (e.g. MetaPhlAn, mOTUs, Bracken) must perform some form of _classification_ prior to profiling.
Not all tools currently have dedicated tips, suggestions and/or recommendations, however we welcome further contributions for existing and additional tools via pull requests to the [nf-core/taxprofiler repository](https://github.com/nf-core/taxprofiler)!
It is unclear whether Bracken is suitable for running long reads, as it makes certain assumptions about read lengths. Furthemore, during testing we found issues where Bracken would fail on the long-read test data.
Therefore currently nf-core/taxprofiler does not run Bracken on data specified as being sequenced with `OXFORD_NANOPORE` in the input samplesheet.
DIAMOND only allows output of a single file format at a time, therefore parameters such `--diamond_save_reads` supplied will result in only aligned reads in SAM format will be produced, no taxonomic profiles will be available. Be aware of this when setting up your pipeline runs, depending on your particular use case.
MALT does not support paired-end reads alignment (unlike other tools), therefore nf-core/taxprofiler aligns these as indepenent files if read-merging is skipped. If you skip merging, you can sum or average the results of the counts of the pairs.
Krona can only be run on MALT output if path to Krona taxonomy database supplied to `--krona_taxonomy_directory`. Therefore if you do not supply the a Krona directory, Krona plots will not be produced for MALT.
> ⚠️ MALT KRONA plots cannot be generated automatically, you must also specify a Krona taxonomy directory with `--krona_taxonomy_directory` if you wish to generate these.
##### Multi-Table Generation
In addition to per-sample profiles, the pipeline also supports generation of 'native' multi-sample taxonomic profiles (i.e., those generated by the taxonomic profiling tools themselves or additional utility scripts provided by the tool authors).
These are executed on a per-database level. I.e., you will get a multi-sample taxon table for each database you provide for each tool and will be placed in the same directory as the directories containing the per-sample profiles.
When you run the above command, Nextflow automatically pulls the pipeline code from GitHub and stores it as a cached version. When running the pipeline after this, it will always use the cached version if available - even if the pipeline has been updated since. To make sure that you're running the latest version of the pipeline, make sure that you regularly update the cached version of the pipeline:
It is a good idea to specify a pipeline version when running the pipeline on your data. This ensures that a specific version of the pipeline code and software are used when you run your pipeline. If you keep using the same tag, you'll be running the same version of the pipeline, even if there have been changes to the code since.
First, go to the [nf-core/taxprofiler releases page](https://github.com/nf-core/taxprofiler/releases) and find the latest pipeline version - numeric only (eg. `1.3.1`). Then specify this when running the pipeline with `-r` (one hyphen) - eg. `-r 1.3.1`. Of course, you can switch to another version by changing the number after the `-r` flag.
This version number will be logged in reports when you run the pipeline, so that you'll know what you used when you look back in the future. For example, at the bottom of the MultiQC reports.
Several generic profiles are bundled with the pipeline which instruct the pipeline to use software packaged using different methods (Docker, Singularity, Podman, Shifter, Charliecloud, Conda) - see below.
> We highly recommend the use of Docker or Singularity containers for full pipeline reproducibility, however when this is not possible, Conda is also supported.
The pipeline also dynamically loads configurations from [https://github.com/nf-core/configs](https://github.com/nf-core/configs) when it runs, making multiple config profiles for various institutional clusters available at run time. For more information and to see if your system is available in these configs please see the [nf-core/configs documentation](https://github.com/nf-core/configs#documentation).
Note that multiple profiles can be loaded, for example: `-profile test,docker` - the order of arguments is important!
They are loaded in sequence, so later profiles can overwrite earlier profiles.
If `-profile` is not specified, the pipeline will run locally and expect all software to be installed and available on the `PATH`. This is _not_ recommended, since it can lead to different results on different machines dependent on the computer enviroment.
- A generic configuration profile to be used with [Docker](https://docker.com/)
-`singularity`
- A generic configuration profile to be used with [Singularity](https://sylabs.io/docs/)
-`podman`
- A generic configuration profile to be used with [Podman](https://podman.io/)
-`shifter`
- A generic configuration profile to be used with [Shifter](https://nersc.gitlab.io/development/shifter/how-to-use/)
-`charliecloud`
- A generic configuration profile to be used with [Charliecloud](https://hpc.github.io/charliecloud/)
-`conda`
- A generic configuration profile to be used with [Conda](https://conda.io/docs/). Please only use Conda as a last resort i.e. when it's not possible to run the pipeline with Docker, Singularity, Podman, Shifter or Charliecloud.
Specify this when restarting a pipeline. Nextflow will use cached results from any pipeline steps where the inputs are the same, continuing from where it got to previously. For input to be considered the same, not only the names must be identical but the files' contents as well. For more info about this parameter, see [this blog post](https://www.nextflow.io/blog/2019/demystifying-nextflow-resume.html).
You can also supply a run name to resume a specific run: `-resume [run-name]`. Use the `nextflow log` command to show previous run names.
### `-c`
Specify the path to a specific config file (this is a core Nextflow command). See the [nf-core website documentation](https://nf-co.re/usage/configuration) for more information.
## Custom configuration
### Resource requests
Whilst the default requirements set within the pipeline will hopefully work for most people and with most input data, you may find that you want to customise the compute resources that the pipeline requests. Each step in the pipeline has a default set of requirements for number of CPUs, memory and time. For most of the steps in the pipeline, if the job exits with any of the error codes specified [here](https://github.com/nf-core/rnaseq/blob/4c27ef5610c87db00c3c5a3eed10b1d161abf575/conf/base.config#L18) it will automatically be resubmitted with higher requests (2 x original, then 3 x original). If it still fails after the third attempt then the pipeline execution is stopped.
For example, if the nf-core/rnaseq pipeline is failing after multiple re-submissions of the `STAR_ALIGN` process due to an exit code of `137` this would indicate that there is an out of memory issue:
[62/149eb0] NOTE: Process `NFCORE_RNASEQ:RNASEQ:ALIGN_STAR:STAR_ALIGN (WT_REP1)` terminated with an error exit status (137) -- Execution is retried (1)
Error executing process > 'NFCORE_RNASEQ:RNASEQ:ALIGN_STAR:STAR_ALIGN (WT_REP1)'
A first step to bypass this error, you could try to increase the amount of CPUs, memory, and time for the whole pipeline. Therefor you can try to increase the resource for the parameters `--max_cpus`, `--max_memory`, and `--max_time`. Based on the error above, you have to increase the amount of memory. Therefore you can go to the [parameter documentation of rnaseq](https://nf-co.re/rnaseq/3.9/parameters) and scroll down to the `show hidden parameter` button to get the default value for `--max_memory`. In this case 128GB, you than can try to run your pipeline again with `--max_memory 200GB -resume` to skip all process, that were already calculated. If you can not increase the resource of the complete pipeline, you can try to adapt the resource for a single process as mentioned below.
To bypass this error you would need to find exactly which resources are set by the `STAR_ALIGN` process. The quickest way is to search for `process STAR_ALIGN` in the [nf-core/rnaseq Github repo](https://github.com/nf-core/rnaseq/search?q=process+STAR_ALIGN).
We have standardised the structure of Nextflow DSL2 pipelines such that all module files will be present in the `modules/` directory and so, based on the search results, the file we want is `modules/nf-core/star/align/main.nf`.
If you click on the link to that file you will notice that there is a `label` directive at the top of the module that is set to [`label process_high`](https://github.com/nf-core/rnaseq/blob/4c27ef5610c87db00c3c5a3eed10b1d161abf575/modules/nf-core/software/star/align/main.nf#L9).
The [Nextflow `label`](https://www.nextflow.io/docs/latest/process.html#label) directive allows us to organise workflow processes in separate groups which can be referenced in a configuration file to select and configure subset of processes having similar computing requirements.
The default values for the `process_high` label are set in the pipeline's [`base.config`](https://github.com/nf-core/rnaseq/blob/4c27ef5610c87db00c3c5a3eed10b1d161abf575/conf/base.config#L33-L37) which in this case is defined as 72GB.
Providing you haven't set any other standard nf-core parameters to **cap** the [maximum resources](https://nf-co.re/usage/configuration#max-resources) used by the pipeline then we can try and bypass the `STAR_ALIGN` process failure by creating a custom config file that sets at least 72GB of memory, in this case increased to 100GB.
The custom config below can then be provided to the pipeline via the [`-c`](#-c) parameter as highlighted in previous sections.
> **NB:** We specify the full process name i.e. `NFCORE_RNASEQ:RNASEQ:ALIGN_STAR:STAR_ALIGN` in the config file because this takes priority over the short name (`STAR_ALIGN`) and allows existing configuration using the full process name to be correctly overridden.
The [Nextflow DSL2](https://www.nextflow.io/docs/latest/dsl2.html) implementation of this pipeline uses one container per process which makes it much easier to maintain and update software dependencies. If for some reason you need to use a different version of a particular tool with the pipeline then you just need to identify the `process` name and override the Nextflow `container` definition for that process using the `withName` declaration. For example, in the [nf-core/viralrecon](https://nf-co.re/viralrecon) pipeline a tool called [Pangolin](https://github.com/cov-lineages/pangolin) has been used during the COVID-19 pandemic to assign lineages to SARS-CoV-2 genome sequenced samples. Given that the lineage assignments change quite frequently it doesn't make sense to re-release the nf-core/viralrecon everytime a new version of Pangolin has been released. However, you can override the default container used by the pipeline by creating a custom config file and passing it as a command-line argument via `-c custom.config`.
1. Check the default version used by the pipeline in the module file for [Pangolin](https://github.com/nf-core/viralrecon/blob/a85d5969f9025409e3618d6c280ef15ce417df65/modules/nf-core/software/pangolin/main.nf#L14-L19)
2. Find the latest version of the Biocontainer available on [Quay.io](https://quay.io/repository/biocontainers/pangolin?tag=latest&tab=tags)
> **NB:** If you wish to periodically update individual tool-specific results (e.g. Pangolin) generated by the pipeline then you must ensure to keep the `work/` directory otherwise the `-resume` ability of the pipeline will be compromised and it will restart from scratch.
### nf-core/configs
In most cases, you will only need to create a custom config as a one-off but if you and others within your organisation are likely to be running nf-core pipelines regularly and need to use the same settings regularly it may be a good idea to request that your custom config file is uploaded to the `nf-core/configs` git repository. Before you do this please can you test that the config file works with your pipeline of choice using the `-c` parameter. You can then create a pull request to the `nf-core/configs` repository with the addition of your config file, associated documentation file (see examples in [`nf-core/configs/docs`](https://github.com/nf-core/configs/tree/master/docs)), and amending [`nfcore_custom.config`](https://github.com/nf-core/configs/blob/master/nfcore_custom.config) to include your custom profile.
See the main [Nextflow documentation](https://www.nextflow.io/docs/latest/config.html) for more information about creating your own configuration files.
If you have any questions or issues please send us a message on [Slack](https://nf-co.re/join/slack) on the [`#configs` channel](https://nfcore.slack.com/channels/configs).
Nextflow handles job submissions and supervises the running jobs. The Nextflow process must run until the pipeline is finished.
The Nextflow `-bg` flag launches Nextflow in the background, detached from your terminal so that the workflow does not stop if you log out of your session. The logs are saved to a file.
Alternatively, you can use `screen` / `tmux` or similar tool to create a detached session which you can log back into at a later time.
Some HPC setups also allow you to run nextflow within a cluster job submitted your job scheduler (from where it submits more jobs).
## Nextflow memory requirements
In some cases, the Nextflow Java virtual machines can start to request a large amount of memory.
We recommend adding the following line to your environment to limit this (typically in `~/.bashrc` or `~./bash_profile`):
Not all taxonomic profilers provide ready-made or default databases. Here we will give brief guidance on how to build custom databases for each supported taxonomic profiler.
You should always consult the documentation of each tool for more information, as here we only provide short minimal-tutorials as quick reference guides (with no guarantee they are up to date).
The following tutorials assumes you already have the tool available (e.g. installed locally, or via conda, docker etc.), and you have already downloaded the FASTA files you wish to build into a database.
Bracken does not require an independent database nor not provide any default databases for classification/profiling, but rather builds upon Kraken2 databases. See [Kraken2](#kraken2-custom-database) for more information on how to build these.
In addition to a Kraken2 database, you also need to have the (average) read lengths (in bp) of your sequencing experiment, the K-mer size used to build the Kraken2 database, and Kraken2 available on your machine.
In total, you need four components: a tab-separated file mapping sequence IDs to taxonomy IDs (`--conversion-table`), a tab-separated file mapping taxonomy IDs to their parents and rank, up to the root of the tree (`--taxonomy-tree`), a pipe-separated file mapping taxonomy IDs to a name (`--name-table`), and the reference sequences.
To build a kaiju database, you need three components: a FASTA file with the protein sequences ,the NCBI taxonomy dump files, and you need to define the uppercase characters of the standard 20 amino acids you wish to include.
> ⚠️ The headers of the protein fasta file must be numeric NCBI taxon identifiers of the protein sequences.
To download the NCBI taxonomy files, please run the following commands:
To build a Kraken2 database you need two components: a taxonomy (consisting of `names.dmp`, `nodes.dmp`, and `*accession2taxid`) files, and the FASTA files you wish to include.
You can follow the Kraken2 [tutorial](https://github.com/DerrickWood/kraken2/blob/master/docs/MANUAL.markdown#custom-databases) for a more detailed description.
First you must make a `seqid2taxid.map` file which is a two column text file containing the FASTA sequence header and the NCBI taxonomy ID for each sequence:
Then make a directory (`<DB_DIR_NAME>/`), containing the `seqid2taxid.map` file, and your FASTA files in a subdirectory called `library/` (these FASTA files can be symlinked). You must then run the `taxonomy` command on the `<DB_DIR_NAME>/` directory, and then build it.
To build a MALT database, you need the FASTA files to include, and an (unzipped) [MEGAN mapping 'db' file](https://software-ab.informatik.uni-tuebingen.de/download/megan6/) for your FASTA type. In addition to the input directory, output directory, and the mapping file database, you also need to specify the sequence type (DNA or Protein) with the `-s` flag.
⚠️ MALT generates very large database files and requires large amounts of RAM. You can reduce both by increasing the step size `-st` (with a reduction in sensitivity).
MetaPhlAn3 does not allow (easy) construction of custom databases. Therefore we recommend to use the prebuilt database of marker genes that is provided by the developers.
To do this you need to have `MetaPhlAn3` installed on your machine.
> 🛈 It is generally not recommended to modify this database yourself, thus this is currently not supported in the pipeline. However, it is possible to customise the existing database by adding your own marker genomes following the instructions [here](https://github.com/biobakery/MetaPhlAn/wiki/MetaPhlAn-3.1#customizing-the-database).
> 🖊️ If using your own database is relevant for you, please contact the nf-core/taxprofiler developers on the [nf-core slack](https://nf-co.re/join) and we will investigate supporting this.
mOTUs does not provide the ability to construct custom databases. Therefore we recommend to use the the prebuilt database of marker genes provided by the developers.
Then supply the `db_mOTU/` path to your nf-core/taxprofiler database input sheet.
> ⚠️ The `db_mOTU/` directory may be downloaded to somewhere in your Python's `site-package` directory. You will have to find this yourself as the exact location varies depends on installation method.
More information on the mOTUs database can be found [here](https://motu-tool.org/installation.html).
