The configuration file config/config.yaml should indicate the path to the tabular file describing the isolates and the raw sequences to be processed.
It is specified as follows:
samples: config/isolates.tsv
The raw sequence files can be freely named and will be standardized later. In the example below, both names are accepted.
isolate forward_file reverse_file
Favorite_isolate /home/rickastley/raw-sequences/Favorite_isolate_R1.fastq.gz /home/rickastley/raw-sequences/Favorite_isolate_R1.fastq.gz
Slowgrower_isolate /home/rickastley/raw-sequences/ComplexName-of-the-isolate_R1.fastq.gz /home/rickastley/raw-sequences/ComplexName-of-the-isolate_R2.fastq.gz
In the configuration directory of the workflow:
cd config/
List all the forward (R1) and reverse (R2) FASTQ files in their directory
ls /data/project_fastq/PROJECT*R1_001.fastq.gz > R1
ls /data/project_fastq/PROJECT*R2_001.fastq.gz > R2
Extract the isolate name located before the first underscore
# Example of the path /data/PROJECT_fastq/PROJECTH117_S106_R1_001.fastq.gz
sed -e 's/^.*PROJECT/PROJECT/' -e 's/_.*//' R1 > sample
Paste together the columns and separate with tabulations (default)
paste -- sample R1 R2 > project
Add the necessary header to generate the required table as project.tsv
cat <(echo -e "isolate\tforward_file\treverse_file" ) project > project.tsv
rm R1 R2 sample project
Note: Don't forget to update the configuration file with the correct filename!
The maximum number of threads to be used for parallelisation should be indicated as follow (and as well during the snakemake invocation with the -c 9 flag):
threads: 9
The CheckM software runs a full tree by default that uses ~40Gb of RAM. If you don't have such resources, use a reduced tree (~14Gb of RAM) by setting the following flag to true:
reduced_tree: false
Place in the resources folder the FASTA files describing the adapters and the phiX sequences and indicate their path with these two flags:
adapters: resources/NexteraPE-PE.fa
phix: resources/phix.fasta
Specify where should the database of bakta should be found or if not where they should be stored using this flag:
bakta_db: /data/bakta_db
Set the minimum contig size to be kept in the assembly in basepair:
min_contig_length: 500