Table of Contents
How to create a workflow¶
CulebrONT allows you to build a workflow using a simple config.yaml
configuration file :
First, provide the data paths
Second, activate the requested tools for assembly and correction.
Third, activate the tools for quality checking of assemblies.
And last, manage the tools parameters.
To create this file, just run:
culebrONT create_config¶
Create config.yaml for run
culebrONT create_config [OPTIONS]
Options
- -c, --configyaml <configyaml>¶
Required Path to create config.yaml
Then, edit the relevant sections of the file to customize your flavor of a workflow.
1. Providing data¶
First, indicate the data path in the config.yaml
configuration file:
DATA:
FASTQ: '/path/to/fastq/directory/'
REF: '/path/to/referencefile.fasta'
GENOME_SIZE: '1m'
FAST5: '/path/to/fast5/directory/'
ILLUMINA: '/path/to/illumina/directory/'
OUTPUT: '/path/to/output/directory/'
Find here a summary table with the description of each data needed to run CulebrONT :
Input |
Description |
---|---|
FASTQ |
Every FASTQ file should contain the whole set of reads to be assembled per individual. Each fastq file will be assembled independently. |
REF |
Only one REFERENCE genome file will be used in each CulebrONT run. This REFERENCE will be used for various quality steps (i.e. ASSEMBLYTICS, QUAST and MAUVE) |
GENOME_SIZE |
Estimated genome size of the assembly can be provided in mega (Mb), giga(Gb) or kilobases (Kb). This size is used by some assemblers (e.g. CANU) and also by the QUAST quality step |
FAST5 |
Nanopolish uses FAST5 files to polish, and Medaka needs FAST5 files if a model training step is requested. Please give the path of the FAST5 folder in the FAST5 DATA parameter. Inside this directory, a subdirectory with the exact same name as the corresponding FASTQ (before the .fastq.gz) is required. For instance, if in the FASTQ directory we have run1.fastq.gz and run2.fastq.gz, CulebrONT is expecting the run1/ and run2/ subdirectories in the FAST5 main directory |
ILLUMINA |
Indicates the path to the directory with Illumina sequence data (in fastq or fastq.gz format), to perform pilon correction, KAT QC or MERQURY QC. Use preferentially paired-end data. All fastq files need to be homogeneous in their extension name. Please use run1_R1 and run1_R2 nomenclature. |
OUTPUT |
output path directory |
Warning
For FASTQ, the naming conventions accepted by CulebrONT are either NAME.fastq.gz or NAME.fq.gz or NAME.fastq or NAME.fq. Use preferentially short names and avoid special characters to avoid report failure. Please do not use the long name provided directly by the sequencing machine.
All fastq files have to be homogeneous on their extension, and can be compressed or not.
Reference fasta file needs a fasta or fa extension, uncompressed.
2. Choose assemblers, polisher and correctors¶
Activate/deactivate assemblers, polishers and correctors as you wish, usin TRUE/FALSE boolean operators. Feel free to activate only assembly, assembly+polishing or assembly+polishing+correction.
Note
If you expect your genome to include a circular replicon (e.g. with prokaryote), it is recommended to activate CIRCULAR steps
Example:
ASSEMBLY:
CANU: true
FLYE: true
MINIASM: false
RAVEN: false
SMARTDENOVO: false
SHASTA: false
POLISHING:
RACON: true
CIRCULAR: false
CORRECTION:
NANOPOLISH: false
MEDAKA: false
PILON: true
3. Choose quality control tools¶
CulebrONT can use several quality control tools to check assemblies.
If BUSCO or QUAST are used, they will run on every fasta assembly generated along the various steps of the pipeline.
If BLOBTOOLS, ASSEMBLYTICS, FLAGSTATS, MERQURY and KAT are activated, only the fasta assembly generated after the last sequence processing step of the pipeline will be checked.
KAT and MERQURY quality tools can be activated but Illumina reads are mandatory in this case. These reads can be compressed or not.
# BUSCO and QUAST will be launched on all activated steps (ASSEMBLY, POLISHING, CORRECTION)
QUALITY:
BUSCO: true
QUAST: true
#### Others quality tools are launched only in last assemblies
BLOBTOOLS: true
ASSEMBLYTICS: true
#### Others quality soft but illumina reads are required
FLAGSTATS: true
If several assemblers are activated, a multiple alignment of the various assemblies for small genomes (<10-20Mbp) can be computed with Mauve.
If you want to improve alignment with MAUVE on circular molecules, it is recommended to activate the Fixstart step.
Only activate MAUVE if you have more than one assembler per sample, more than one quality step and small genomes.
#### Alignment of the various assemblies derived from a fastq file for small genomes (<10-20Mbp);
MSA:
4. Parameters for some specific tools¶
You can manage tools parameters on the params section in the config.yaml
file.
Racon` specific options:
Racon can be launched recursively from 1 to 9 rounds.
Medaka
specific options:
If ‘SEGMENTATION’ is false, there is no segmentation of contigs for medaka
If ‘MEDAKA_TRAIN_WITH_REF’ is activated, Medaka launchs the training using the reference found in ‘DATA/REF’ path parameter. Medaka will then not take into account other Medaka model parameters and will use the resulting trained model instead.
If ‘MEDAKA_TRAIN_WITH_REF’ is deactivated, Medaka does not launch training, but uses instead the model provided in ‘MEDAKA_MODEL_PATH’ parameter. Give to CulebrONT the path of the Medaka model OR just the model name in order to correct assemblies. This parameter could not be empty.
Important
Medaka models can be downloaded from the Medaka repository. You need to install git lfs
(see documentation here https://git-lfs.github.com/) to download largest files before git clone https://github.com/nanoporetech/medaka.git\
.
Pilon
specific options:
We fixed the java memory parameter in the Singularity.culebront_tools to 8G. If you need to allocate more memory, change this value using
sed -i "s/-Xmx1g/-Xmx8g/g" /usr/local/miniconda/miniconda3/envs/pilon/bin/pilon
in theContainers/Singularity.culebront_tools.def
recipe file before building the Singularity image.
Busco
specific options:
If BUSCO is activated, you must provide to CulebrONT the path of a Busco database OR only the database name (See the Busco documentation).This parameter cannot be empty.
Blobtools
specific options:
* Nodes and names from the NCBI taxdump database can be download here : https://github.com/DRL/blobtools#download-ncbi-taxdump-and-create-nodesdb
The standard parameters used in CulebrONT are shown below. Feel free to adapt it to your own requirements.
############ PARAMS ################
params:
#### ASSEMBLY
MINIMAP2:
PRESET_OPTION: 'map-ont' # -x minimap2 preset option is map-pb by default (map-pb, map-ont etc)
FLYE:
MODE : '--nano-raw'
OPTIONS: '' ## use --scaffold if flye>=2.9 # you can also use --resume option
CANU:
MODE : '-nanopore'
OPTIONS: 'useGrid=false'
SMARTDENOVO:
KMER_SIZE: 16
OPTIONS: '-J 5000'
SHASTA:
MEM_MODE: 'filesystem'
MEM_BACKING: 'disk'
OPTIONS: '--Reads.minReadLength 0'
#### CIRCULAR
CIRCLATOR:
OPTIONS: ''
#### POLISHING
RACON:
RACON_ROUNDS: 2 #1 to 9
#### CORRECTION
CORRECTION_SEGMENTATION:
SEGMENT_LEN: 50000 # segment length to split assembly and correct it default=50000
OVERLAP_LEN: 200 # overlap length between segments default=200
NANOPOLISH:
OPTIONS: ''
MEDAKA:
SEGMENTATION: true # if false, there is no segmentation of contigs for medaka
MEDAKA_TRAIN_WITH_REF: false # if 'MEDAKA_TRAIN_WITH_REF' is True, training uses reference to found in DATA REF param.
# Medaka does not take in count other parameters below if MEDAKA_TRAIN_WITH_REF is TRUE.
MEDAKA_MODEL_PATH: 'r941_min_high_g303' # use a path if you have downloaded a model (or you want to use your own trained model) OR a simple string like 'r941_min_high_g303'
MEDAKA_FEATURES_OPTIONS: '--batch_size 10 --chunk_len 100 --chunk_ovlp 10'
MEDAKA_TRAIN_OPTIONS: '--batch_size 10 --epochs 500 '
MEDAKA_CONSENSUS_OPTIONS: '--batch 200 '
PILON:
PILON_ROUNDS: 2 #1 to 9
OPTIONS: ''
#### QUALITY
BUSCO:
#DATABASE: "DATA_DIR/Data-Xoo-sub/bacteria_odb10"
DATABASE: 'bacteria_odb10 --update-data ' # use a path if you have downloaded a taxonomic database from busco OR a simple string like 'bacteria_odb10'
MODEL: 'genome'
SP: '' #--augustus-specie parameter on busco
QUAST:
GFF: ''
OPTIONS: '--large'
DIAMOND:
DATABASE: 'DATA_DIR/Data-Xoo-sub/testBacteria.dmnd'
MUMMER:
MINMATCH: 100 # is -l option with default 20 on MUMMER
MINCLUSTER: 500 # is -c option with default 65 on MUMMER
ASSEMBLYTICS:
UNIQUE_ANCHOR_LEN: 10000
MIN_VARIANT_SIZE: 50
MAX_VARIANT_SIZE: 10000
BLOBTOOLS:
NAMES: 'DATA_DIR/Data-Xoo-sub/blobtools/names.dmp'
NODES: 'DATA_DIR/Data-Xoo-sub/blobtools/nodes.dmp'
Warning
Please check documentation of each tool (outside of CulebrONT, and make sure that the settings are correct!)
How to run the workflow¶
Before attempting to run CulebrONT, please verify that you have already modified the config.yaml
file as explained in 1. Providing data.
If you installed CulebrONT on a HPC cluster with a job scheduler, you can run:
culebrONT run_cluster¶
- Example:
culebrONT run_cluster -c config.yaml –dry-run –jobs 200
culebrONT run_cluster [OPTIONS] [SNAKEMAKE_OTHER]...
Options
- -c, --config <config>¶
Required Configuration file for run culebrONT
- -pdf, --pdf¶
Run snakemake with –dag, –rulegraph and –filegraph
- Default
False
Arguments
- SNAKEMAKE_OTHER¶
Optional argument(s)
culebrONT run_local¶
Example:
culebrONT run_local -c config.yaml –threads 8 –dry-run
culebrONT run_local -c config.yaml –threads 8 –singularity-args ‘–bind /mnt:/mnt’
# in LOCAL using 6 threads for Canu assembly from the total 8 threads
culebrONT run_local -c config.yaml –threads 8 –set-threads run_canu=6
culebrONT run_local [OPTIONS] [SNAKEMAKE_OTHER]...
Options
- -c, --config <config>¶
Required Configuration file for run culebrONT
- -t, --threads <threads>¶
Required Number of threads
- -p, --pdf¶
Run snakemake with –dag, –rulegraph and –filegraph
Arguments
- SNAKEMAKE_OTHER¶
Optional argument(s)
Advance run¶
Providing more resources¶
If the cluster default resources are not sufficient, you can edit the cluster_config.yaml
file. See 2. Adapting cluster_config.yaml:
culebrONT edit_cluster_config¶
Edit cluster_config.yaml use by profile
culebrONT edit_cluster_config [OPTIONS]
Providing your own tools_config.yaml¶
To change the tools used in a CulebrONT workflow, you can see 3. How to configure tools_path.yaml
culebrONT edit_tools¶
Edit own tools version
culebrONT edit_tools [OPTIONS]
Options
- -r, --restore¶
Restore default tools_config.yaml (from install)
- Default
False
Output on CulebrONT¶
The architecture of the CulebrONT output is designed as follow:
OUTPUT_CULEBRONT_CIRCULAR/
├── SAMPLE-1
│ ├── AGGREGATED_QC
│ │ ├── DATA
│ │ ├── MAUVE_ALIGN
│ │ └── QUAST_RESULTS
│ ├── ASSEMBLERS
│ │ ├── CANU
│ │ │ ├── ASSEMBLER
│ │ │ ├── CORRECTION
│ │ │ ├── FIXSTART
│ │ │ ├── POLISHING
│ │ │ └── QUALITY
│ │ ├── FLYE
│ │ │ ├── ...
│ │ ├── MINIASM
│ │ │ ├── ...
│ │ ├── RAVEN
│ │ │ ├── ...
│ │ ├── SHASTA
│ │ │ ├── ...
│ │ └── SMARTDENOVO
│ │ │ ├── ...
│ ├── DIVERS
│ │ └── FASTQ2FASTA
│ ├── LOGS
│ └── REPORT
└── FINAL_REPORT
├── SAMPLE-2 ...
Report¶
CulebrONT generates a useful HTML report, including the versions of tools used and, for each fastq, a summary of statistics. Please have a look at example … and enjoy !!
Note
Because of constraints imposed by Snakemake, we cannot include the version of bwa and seqtk in the report https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#step-5-loggin. If you want to know the versions of these tools, go check by yourself ^^.
Important
To visualise the report created by CulebrONT, transfer the folder FINAL_RESULTS
on your local computer and open it on any web browser.