KneadData

You can install KneadData from a cloned repository:

1. Clone the repository: $ git clone https://github.com/biobakery/kneaddata.git
2. Navigate to the directory: $ cd kneaddata
3. Run the installation: $ python setup.py install

Options:

• Bypass dependencies: Add --bypass-dependencies-install to skip Trimmomatic and Bowtie2 installation.
• User-level install: Add --user if you lack write permissions to /usr/lib/. Remember to add $HOME/.local/bin/ to your $PATH if necessary.

The recommended way to install KneadData is using pip. This method automatically installs Trimmomatic and Bowtie2.

• Standard install: pip install kneaddata
• Bypass dependency installation: Use --install-option='--bypass-dependencies-install' if you want to skip the automatic installation of Trimmomatic and Bowtie2.
• User-level install: If you lack write permissions to /usr/lib/, use the --user flag. This installs the package into $HOME/.local.

Important: If you use --user, you may need to manually add $HOME/.local/bin/ to your $PATH if you encounter the error kneaddata: command not found.

$ pip install kneaddata
# Or for user-level installation
$ pip install --user kneaddata

You can use the kneaddata_database utility to download pre-indexed common databases. To download the indexed human reference database, use the --download flag with human and specify the alignment tool (e.g., bowtie2).

Replace $DIR with the full path to the directory where you want to store the database.

$ kneaddata_database --download human bowtie2 $DIR

To create a custom Bowtie2 database, use the bowtie2-build command. This requires a reference FASTA file and a name for the database.

Note for SILVA ribosomal_RNA: If you are creating a database from SILVA ribosomal RNA sequences, you must first run the modify_RNA_to_DNA.py script to convert 'U's to 'T's in the FASTA sequences before running bowtie2-build.

# Standard Bowtie2 build
$ bowtie2-build <reference> <db-name>

# Required preprocessing for SILVA ribosomal_RNA
$ python -u modify_RNA_to_DNA.py input.fasta output.fa
$ bowtie2-build output.fa <db-name>

KneadData provides pre-built Bowtie2 databases for contamination removal and ribosomal RNA filtering. You can download the following versions:

• human_genome (v0.1)
• human_transcriptome (v0.1)
• mouse_C57BL (v0.1)
• ribosomal_RNA (v0.2)

These databases are typically used to identify and filter out host or unwanted sequences from sequencing data.

When using the --bypass-trim flag, KneadData expects input files to follow its post-trim naming convention (e.g., *.trimmed.fastq). If you provide a standard input.fastq, the run may crash with an unclear error.

Workaround: Rename your input file to include the .trimmed suffix before running KneadData.

mv input.fastq input.trimmed.fastq

You can perform a test run using the provided demo files. This example uses an unpaired FASTQ file and a demo database to generate output files including trimmed reads, contaminant reads, and logs.

$ kneaddata --unpaired examples/demo.fastq --reference-db examples/demo_db --output kneaddata_demo_output

To generate a ribosomal_RNA Bowtie2 database, you must first convert the RNA fasta to DNA format using the provided script, then proceed with synthetic read generation and index building:

1. Convert RNA to DNA using modify_RNA_to_DNA.py.
2. Extract a subset of the fasta.
3. Generate synthetic reads using art_illumina.
4. Build the Bowtie2 index.

# 1. Convert RNA to DNA
$ python -u modify_RNA_to_DNA.py ribosomal_RNA_initial.fasta  ribosomal_RNA.fasta

# 2. Extract genome subset
$ bowtie2-inspect /ribosomal_RNA | head -n 1000 > ribosomal_RNA.fasta

# 3. Generate synthetic reads
$ ./art_illumina -i ribosomal_RNA.fasta -l 100 -o ribosomal_RNA_synthetic_reads -sam -ss HS20 -f 0.3 --id ribosomal_RNA

# 4. Build Bowtie2 database
$ bowtie2-build ribosomal_RNA.fasta ribosomal_RNA

To manually generate Bowtie2 databases and synthetic reads for human_genome, human_transcriptome, or mouse_C57BL, follow these steps using bowtie2-inspect, art_illumina, and bowtie2-build:

1. Extract a subset of the genome fasta.
2. Generate synthetic reads using art_illumina.
3. Build the Bowtie2 index.

# 1. Extract genome subset
$ bowtie2-inspect /genome_demo | head -n 1000 > genome_demo.fasta

# 2. Generate synthetic reads
$ ./art_illumina -i genome_demo.fasta -l 100 -o genome_demo_synthetic_reads -sam -ss HS20 -f 0.3 --id genome_demo

# 3. Build Bowtie2 database
$ bowtie2-build genome_demo.fasta genome_demo