Naming conventions

The key words “MUST”, “MUST NOT”, “SHOULD”, etc. are to be interpreted as described in RFC 2119.

Name format of subworkflow files

Subworkflow names SHOULD be based on their module composition.

For short chains of modules without conditional logic, use the format <file_type>_<operation_1>_<operation_n>_<tool_1>_<tool_n>. For example, bam_sort_stats_samtools where bam = <file_type>, sort = <operation> and samtools = <tool>. Exclude routine operations from the name (for example, indexing after BAM creation). Collapse directly related steps into a general name. For example, if a binning tool has three required steps (<tool> split, <tool> calculate, <tool> merge) to perform contig binning, collapse these into one (for example, fasta_binning_concoct, rather than fasta_split_calculate_merge_concoct).

The above naming scheme may not be appropriate if:

• The subworkflow has a large number of steps (excluding routine operations)
• The sequence of steps differs based on input arguments
• The module complement is likely to change over time

In these cases, name the subworkflow according to its purpose and logical operations rather than the module complement. For example, a subworkflow that takes FASTQ files, performs multiple QC checks, applies a trimming operation, filters and sets strandedness would be named fastq_qc_trim_filter_setstrandedness. This tells users the input and logical steps without including every conditional or module in the name.

Use all lowercase for the directory structure. For example, subworkflows/nf-core/bam_sort_stats_samtools/.

For assistance with naming subworkflows, particularly complex subworkflows, post on the nf-core Slack #subworkflows channel (you can join with this invite) to discuss options.

Name format of subworkflow parameters

Parameter names MUST use the snake_case convention.

Name format subworkflow functions

Function names MUST use the camelCase convention.

Name format subworkflow channels

Channel names MUST use the snake_case convention and be lowercase.

Input channel name structure

Input channel names SHOULD signify the input object type. Prefix single value input channels with val_. Prefix input channels with multiple elements (e.g., meta map and file) with ch_.

Output channel name structure

Output channel names SHOULD be based on the major output file of that channel. For example, emit an output channel of [[meta], bam] as bam, not ch_bam. This provides more intuitive use of these output objects downstream with the .out attribute.