nf-core/deepmutscan

Edit

Introduction

nf-core/deepmutscan is a workflow for the analysis of deep mutational scanning (DMS) data. It takes short-read next-generation DNA sequencing data of a site-saturation mutagenesis library – typically before (input) and after (output) a variant selection experiment – and turns them into quality-controlled, error-corrected variant frequencies and, optionally, relative variant fitness estimates.

The pipeline was mainly designed for shotgun sequencing of long open reading frames (ORFs), where the mutated gene is randomly fragmented before sequencing and where resulting variant count signals are sparse, but it processes amplicon (tile) sequencing data in the same way.

This page explains how to prepare the inputs, how to run the pipeline, and what each processing step does and why. It is intended both for new DMS users and for developers who want to understand the exact design choices.

Samplesheet input

You will need to create a samplesheet with information about the sequencing libraries you would like to analyse before running the pipeline. Use the --input parameter to specify its location. It has to be a comma-separated file with 5 columns and a header row, as shown in the example below.

--input '[path to samplesheet file]'

sample,type,replicate,file1,file2
ORF1,input,1,/reads/input1_R1.fastq.gz,/reads/input1_R2.fastq.gz
ORF1,input,2,/reads/input2_R1.fastq.gz,/reads/input2_R2.fastq.gz
ORF1,output,1,/reads/output1_R1.fastq.gz,/reads/output1_R2.fastq.gz
ORF1,output,2,/reads/output2_R1.fastq.gz,/reads/output2_R2.fastq.gz

Each library gets an internal identifier <sample>_<type>_<replicate>_pe, which is used to name its output files and folders.

Reference and reading frame

As of v1.0.0, all libraries in one run of nf-core/deepmutscan must stem from the same gene. Provide its wildtype sequence as a FASTA file via --fasta. We recommend that the reference contains N- and C-terminal flanking sequences of approximately one read length (e.g. from surrounding vector regions amplified during the sequencing library preparation), which helps to align reads more uniformly at the ends of the ORF.

Use --reading_frame to give the position of the mutagenised ORF within that sequence, as 1-based, inclusive nucleotide coordinates in the format start-stop: start is the first nucleotide of the first codon and stop the last nucleotide of the last codon to be analysed, so the length of the range must be divisible by three. For example, --reading_frame 151-450 analyses exactly 100 codons of the reference FASTA.

Use --mutagenesis_type to define which codons were programmed at each position (default nnk). This determines which observed variants count as library variants for filtering, QC and fitness estimation:

• nnk, nns, nnh, nnn: degenerate codons, as used e.g. in nicking mutagenesis
• nnk_nns, nnk_nns_nnh: position-dependent degenerate codons, chosen by the third base of the wildtype codon: NNS where it is T, NNH where it is G (nnk_nns_nnh only), and NNK otherwise
• custom: a user-defined codon library, provided with --custom_codon_library as a .csv file. Either give one global comma-separated list of codons without a header (e.g. AAA,AAC,AAG), or a position-wise list: a header line containing the word Position, followed by one row per codon position with the position and its allowed codons (e.g. 1,ACG,AAA,ACA).

Running the pipeline

The typical command for running the pipeline (here on a hypothetical gene of exactly 100 codons) is as follows:

nextflow run nf-core/deepmutscan \
   -profile <docker/singularity/.../institute> \
   --input ./samplesheet.csv \
   --fasta ./ref.fa \
   --reading_frame 151-450 \
   --outdir ./results

This will launch the pipeline with the docker configuration profile (or the profile of your choice, see below). It performs the default read QC, alignment, filtering, merging, variant counting, sequencing-error correction and library QC for every library in the samplesheet. Add --fitness to also estimate variant fitness from matched input and output libraries, and --dimsum and/or --mutscan for the additional fitness estimators.

Note that the pipeline will create the following files in your working directory:

work                # Directory containing the nextflow working files
<OUTDIR>            # Finished results in specified location (defined with --outdir)
.nextflow_log       # Log file from Nextflow
# Other nextflow hidden files, eg. history of pipeline runs and old logs.

If you wish to repeatedly use the same parameters for multiple runs, rather than specifying each flag in the command, you can specify these in a params file.

Pipeline settings can be provided in a yaml or json file via -params-file <file>.

The above pipeline run specified with a params file in yaml format:

nextflow run nf-core/deepmutscan -profile docker -params-file params.yaml

with:

input: './samplesheet.csv'
outdir: './results/'
fasta: './ref.fa'
reading_frame: '151-450'
fitness: true
<...>

You can also generate such YAML/JSON files via nf-core/launch.

Pipeline steps in detail

1. Alignment

All paired-end raw reads are first quality-checked with FastQC and then aligned to the reference with BWA-MEM. BWA-MEM is a highly efficient mapping algorithm for reads ≥ 70 bp.

2. Filtering

Alignments are filtered with samtools view: only mapped, non-secondary alignments with a mapping quality ≥ 30 and without insertions, deletions or splice gaps are kept before read merging, as indel-containing reads are most likely artefacts in substitution libraries.

For long ORF site-saturation mutagenesis libraries, most aligned shotgun sequencing reads match the reference exactly. These wildtype reads are retained: they are ignored by the variant counter but contribute to the overall sequencing coverage and sequencing-error estimation.

3. Read merging

Even the highest-quality sequencing platforms do not have perfect base accuracy. To minimise the effect of base errors, which would otherwise be counted as false mutations, nf-core/deepmutscan uses the overlap of each read pair. As base errors on the forward and reverse read are largely independent, vsearch –fastq_mergepairs converts each read pair into a single consensus read with adjusted base qualities. Read pairs that cannot be merged (overlap shorter than 10 bp), and reads whose mate was removed during filtering, are discarded. Merged reads are then re-aligned with BWA-MEM, coordinate-sorted and indexed with samtools.

4. Variant counting

Merged reads are screened for base-level mismatches against the reference. nf-core/deepmutscan uses a light-weight Python counter (built on pysam and polars) that counts all single, double, triple and higher-order nucleotide changes per read, together with the sequencing coverage of every position. This replaces the GATK AnalyzeSaturationMutagenesis tool used in earlier versions; a description of every column is written next to each count table (variant_counts_columns.tsv).

Two parameters tune the counter:

• --base_qual (default 40): minimum Phred base quality for a mismatch to be counted, and for a base to contribute to coverage.
• --min_flank (default 2): minimum distance (in nucleotides) of a mismatch or covered base from the read end. Set to 0 to disable.

5. Library annotation and filtering

Variant counts are annotated with their codon and amino acid changes within --reading_frame, and filtered for the variants that were programmed by the chosen --mutagenesis_type. In addition, a library-completed table lists every programmed variant, including those that were not observed (with a near-zero placeholder count), for the count distribution plots.

6. Single-nucleotide variant error correction

Sequencing errors occur at typically low, but position- and substitution-specific rates. With very deep sequencing, they inflate the counts of variants that differ from the wildtype by only a single nucleotide. Variants with two or three nucleotide changes in one codon, however, are hardly affected. Select a correction strategy with --error_correction:

• false_doubles (default): the library only contains single-codon changes, so reads carrying a programmed multi-nucleotide codon variant plus an unprogrammed single-nucleotide change in a nearby codon (“false double mutants”) reveal the sequencing error rate of that single-nucleotide change. The error rate is estimated for every single-nucleotide variant from these read-linked observations within --false_doubles_codon_window codons (default 40), and the expected number of erroneous counts is subtracted. No extra data are required. If a library contains too few false double mutants to model their coverage along the read (e.g. very small or shallow libraries), no correction is applied and a warning is written to the log. Two estimators are available via --false_doubles_method:
  • mle (default): a per-variant maximum-likelihood error rate.
  • eb: an empirical-Bayes estimate that shrinks the per-variant rate towards the mean of its substitution class. It is steadier for sparsely observed variants, e.g. in shallowly sequenced libraries.
• wildtype: subtracts the variant-specific background measured by additional deep sequencing of the unmutated template from every library variant. Add these libraries to the samplesheet with type set to wildtype and the same sample name as the libraries they should correct. The pipeline stops with an error if a sample has no wildtype library. In this mode, the wildtype libraries themselves are not included in the library QC and fitness outputs.
• none: no correction; counts are used as observed.

Corrected counts are used for every count-dependent step downstream (count heatmaps, positional QC and fitness). The uncorrected tables are kept next to the corrected ones, and an interactive HTML report shows the effect of the correction across all libraries of the run.

7. DMS library quality control

Based on the library-filtered (and error-corrected) counts, nf-core/deepmutscan produces per-library visualisations to assess mutation efficiency, coverage and saturation along the ORF:

• heatmaps of variant counts and counts per coverage, per position and amino acid
• sorted, log-scale distributions of counts per coverage of all programmed variants, also stratified by the number of changed nucleotides, summarised by the log10 ratio of their 90th and 10th percentiles
• sliding-window profiles of coverage and counts along the ORF (window size set by --sliding_window_size, default 10 codons), with a reference line for the coverage needed to observe every amino acid variant --aimed_cov times (default 100)
• a sequencing-depth rarefaction curve that estimates the fraction of programmed variants recovered at lower sequencing depths (--run_seqdepth, on by default)

8. Fitness estimation (optional)

With --fitness, libraries of each sample are grouped into a merged count table of all input and output replicates. Fitness is then estimated per replicate as the natural logarithm of each variant’s output-to-input count ratio, relative to that of a synonymous wildtype proxy variant. As the proxy, the pipeline picks the synonymous variant with two nucleotide changes in one codon (or, if there is none, one nucleotide change) with the highest mean input count. Non-synonymous variants encoding the same amino acid sequence are aggregated. Per-replicate fitness values are linearly rescaled so that the median of synonymous variants is 0 and the median of stop codon variants is -1 (if one of these anchors is missing, fitness values are only centred), and summarised as the mean and, with more than one replicate, the standard deviation across replicates.

Two parameters control which variants are scored and how dropouts are treated:

• --min_counts (default 10): minimum number of reads a variant needs in every input replicate to receive a fitness estimate. Variants with fewer input reads are poorly measured and are excluded.
• --output_pseudocount (default 1): pseudocount added to output replicates in which a variant observed in the input was not detected (“dropouts”), so that strongly depleted variants still receive a finite fitness value. Set it to 0 to disable; dropout variants then receive no fitness estimate in that replicate.

Two established statistical frameworks can be run on the same merged counts:

• --dimsum: DiMSum fitness estimates (stages 4–5 of DiMSum, without its own read processing). --min_counts and --output_pseudocount are passed on to DiMSum (--fitnessMinInputCountAll, --fitnessDropoutPseudocount).
• --mutscan: mutscan enrichment estimates with edgeR and limma. Variants below --min_counts are removed before the analysis.

Interactive variant effect inspection tool (--pdb)

When --fitness is set and a wildtype 3D structure is supplied via --pdb <structure.pdb> (an experimental structure, an AlphaFold DB model, or any PDB file matching the wildtype ORF), the pipeline additionally builds a portable, interactive HTML tool. It projects per-residue fitness, coverage, counts, counts per coverage and – when error correction is on – the positional error bias onto the structure; clicking a residue shows the effect of each substitution. Structure prediction is not part of the v1.0.0 release, so a user-specified structure file is required to enable the tool.

9. Reporting

nf-core/deepmutscan writes a single, self-contained deepmutscan_report.html to the top of the results directory. It embeds the library QC plots, the error-correction report, the fitness results (and DiMSum/mutscan output, if run), the MultiQC report, run statistics and the citations of all tools that were used, so it can be shared as a single file.

Updating the pipeline

When you run the original command above, 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 this cached version if available - even if the pipeline has been updated since. To make sure that you are running the latest version of the pipeline, make sure that you regularly update the cached version:

nextflow pull nf-core/deepmutscan

Reproducibility

It is a good idea to specify the pipeline version when running the pipeline on your data. This ensures that a specific version of the pipeline code and software are used consistently. 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/deepmutscan releases page and find the latest pipeline version - numeric only (eg. 1.0.0). Then specify this when running the pipeline with -r (one hyphen) - eg. -r 1.0.0.

This version number will also be logged in reports when you run the pipeline, for example at the bottom of the MultiQC reports.

To further assist in reproducibility, you can share and reuse parameter files to repeat pipeline runs with the same settings without having to write out a command with every single parameter.

Core Nextflow arguments

-profile

Use this parameter to choose a configuration profile. Profiles can give configuration presets for different compute environments.

Several generic profiles are bundled with the pipeline which instruct the pipeline to use software packaged using different methods (Docker, Singularity, Podman, Shifter, Charliecloud, Apptainer, Conda) - see below.

The pipeline also dynamically loads configurations from 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 check if your system is supported, please see the 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 may lead to varying or even irreproducible results across users’ different computer environments.

• test
  • A profile with a complete configuration for automated testing
  • Includes links to test data so needs no other parameters
• docker
  • A generic configuration profile to be used with Docker
• singularity
  • A generic configuration profile to be used with Singularity
• podman
  • A generic configuration profile to be used with Podman
• shifter
  • A generic configuration profile to be used with Shifter
• charliecloud
  • A generic configuration profile to be used with Charliecloud
• apptainer
  • A generic configuration profile to be used with Apptainer
• wave
  • A generic configuration profile to enable Wave containers. Use together with one of the above (requires Nextflow 24.03.0-edge or later).
• conda
  • A generic configuration profile to be used with Conda. Please only use Conda as a last resort i.e. when it’s not possible to run the pipeline with Docker, Singularity, Podman, Shifter, Charliecloud, or Apptainer.

-resume

Specify this when restarting a pipeline. Nextflow will use cached results (from within the /work directory) 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.

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 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 pipeline steps, if the job exits with any of the error codes specified here it will automatically be resubmitted with higher resources request (2 x original, then 3 x original). If it still fails after the third attempt then the pipeline execution is stopped.

To change the resource requests, please see the max resources and customise process resources section of the nf-core website.

Custom Containers

In some cases, you may wish to change the container or conda environment used by a pipeline steps for a particular tool. By default, nf-core pipelines use containers and software from the biocontainers or bioconda projects. However, in some cases the pipeline specified version maybe out of date.

To use a different container from the default container or conda environment specified in a pipeline, please see the updating tool versions section of the nf-core website.

Custom Tool Arguments

A pipeline might not always support every possible argument or option of a particular tool used in pipeline. Fortunately, nf-core pipelines provide some freedom to users to insert additional parameters that the pipeline does not include by default.

To learn how to provide additional arguments to a particular tool of the pipeline, please see the customising tool arguments section of the nf-core website.

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), and amending nfcore_custom.config to include your custom profile.

See the main Nextflow documentation for more information about creating your own configuration files.

If you have any questions or issues please send us a message on Slack on the #configs channel.

Running in the background

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):

NXF_OPTS='-Xms1g -Xmx4g'