Usage

Samplesheet

The samplesheet is a CSV file which contains information about the samples to be analysed in the pipeline. A header is required.

modBAM input (e.g. Dorado)BAM + CSV input (e.g. m6Anet)

This is suitable for modifications analysed by Dorado.

samplesheet.csv

name,group,path_modbam
sample1,group1,/base_and_modcalled/reads1.bam
sample2,group1,/base_and_modcalled/reads2.bam
sample3,group1,/base_and_modcalled/reads3.bam
sample4,group2,/base_and_modcalled/reads4.bam
sample5,group2,/base_and_modcalled/reads5.bam
sample6,group2,/base_and_modcalled/reads6.bam

name: a unique name for each sample
group: the experimental group or condition for each sample
path_modbam: path to pre-basecalled Dorado modification data for each sample.
The file should be a .bam file in 'modbam' format i.e. with tags MM and ML. See the Dorado documentation for more information.

Start Mako:

bash

export REFERENCE_GENOME=/ref/GRCh38.p14.genome.fa \
       REFERENCE_TXOME=/ref/gencode.v50.transcripts.fa \
       REFERENCE_GTF=/ref/gencode.v50.annotation.gtf \
       OUTDIR=results

# run Mako
nextflow run main.nf \
  -profile singularity \
  --samplesheet samplesheet.csv \
  --dataset_name test_dataset \
  --outdir $OUTDIR \
  --input_format modbam \
  --mod_threshold 0.5 \
  --genome $REFERENCE_GENOME \
  --transcriptome $REFERENCE_TXOME \
  --gtf $REFERENCE_GTF

# launch Makoview
./$OUTDIR/makoview/launch_makoview.sh

This is suitable for modifications analysed by m6Anet.

samplesheet.csv

name,group,path_bam,path_csv
sample1,group1,/basecalled/reads1.bam,/m6anet_out_1/data.indiv_proba.csv
sample2,group1,/basecalled/reads2.bam,/m6anet_out_2/data.indiv_proba.csv
sample3,group1,/basecalled/reads3.bam,/m6anet_out_3/data.indiv_proba.csv
sample4,group2,/basecalled/reads4.bam,/m6anet_out_4/data.indiv_proba.csv
sample5,group2,/basecalled/reads5.bam,/m6anet_out_5/data.indiv_proba.csv
sample6,group2,/basecalled/reads6.bam,/m6anet_out_6/data.indiv_proba.csv

name: a unique name for each sample
group: the experimental group or condition for each sample
path_bam: path to pre-basecalled and aligned reads in BAM format
path_csv: path to a CSV table containing rows with each analysed modification probability. If using m6Anet, this file is data.indiv_proba.csv. The file must have columns transcript_id, transcript_position, read_index, probability_modified.

Start Mako:

bash

export REFERENCE_GENOME=/ref/GRCh38.p14.genome.fa \
       REFERENCE_TXOME=/ref/gencode.v50.transcripts.fa \
       REFERENCE_GTF=/ref/gencode.v50.annotation.gtf \
       OUTDIR=results

# run Mako
nextflow run main.nf \
    -profile singularity \
    --samplesheet samplesheet.csv \
    --dataset_name test_dataset \
    --outdir $OUTDIR \
    --input_format table \
    --mod_threshold 0.033379376 \
    --genome $REFERENCE_GENOME \
    --transcriptome $REFERENCE_TXOME \
    --gtf $REFERENCE_GTF

# launch Makoview
./$OUTDIR/makoview/launch_makoview.sh

Modification thresholds for m6Anet

In this snippet, the modification probability threshold 0.033379376 was the recommended value from m6Anet documentation for their HCT116_RNA002 (human) model (beware m6Anet offers multiple models). Suitable threshold values may differ greatly between tools/models, so we suggest always checking recommendations of tools/models used.

Two groups should be provided to identify differential modifications between conditions. Group names should be alphanumeric and without spaces. The underlying models take the first group alphabetically as the reference level, and the second group alphabetically as the treatment level.

Parameters

Note

If your institution has an nf-core configuration available, you can access it through -profile i.e. -profile wehi to use the WEHI Milton HPC. See Execution Profiles for more.

Dependencies for Mako are distributed as container images for Docker and Singularity (Apptainer). If not using an nf-core configuration, run Mako using -profile docker or -profile singularity.

Input/output options

Define where the pipeline should find input data and save output data.

Parameter	Description	Default
`dataset_name`	The name of this dataset, used in output files and within visualisations
`samplesheet`	Path to comma-separated file containing information about the samples in the experiment.
`outdir`	The output directory where the results will be saved. You have to use absolute paths to storage on Cloud infrastructure.
`publish_dir_mode`	How to copy output from the work directory to the output directory (accepted: `copy`\|`symlink`)	copy
`input_format`	Whether the input format is 'modBAM' or table (TSV) form. (accepted: `modbam`\|`table`)

Pipeline parameters

Parameter	Description	Default
`genome`	Reference genome, in .fasta format
`transcriptome`	Reference transcriptome, in .fasta format
`gtf`	GTF annotation file

Modification calling parameters

Settings which control which sites are selected for differential calling.

Parameter	Description	Default
`method`	Model to use to call differentially modified sites (accepted: `adaptive`\|`binomial`\|`beta_binomial`\|`dss`)	adaptive
`mod_threshold`	A passing threshold to determine a base modification.	0.5
`mod_filter`	Reads with a modification probability within this interval will be filtered out and not used in the Dorado model. The format should be `a,b` i.e. `0.2,0.8`. By default, no filtering is done.
`min_reads_per_sample`	How many reads each sample should have at a site in order for the site to be considered.	5

Other parameters

Parameter	Description	Default
`testing`	Only activated during stub runs
`container_registry`	Container registry domain/path
`container_tag`	Container image tag