SyFi

Pipeline

SyFi is divided into three sequential modules:

• Main: This pipeline uses Illumina reads, contigs and a sequence target (e.g., 16S) to obtain the target haplotypes abundances ratio.
• Amplicon: This pipeline retrieves amplicon fingerprints from the gene fingerprints from the first module using in silico primers.
• Quant: This pipeline takes the results from the first module and quantify the fingerprint abundance from amplicon sequencing data

Dependencies

The pipeline depends on:

• Blast
• Bwa-Mem2
• Samtools
• GATK
• Picard
• BCFtools
• Whatshap
• SeqKit
• Seqtk
• Kallisto
• Bedtools
• R
• Salmon
• Qiime2

Installation

Build conda environment

Conda:

The conda environment is supplemented in the repository. You can create the environment using mamba env create -f SyFi.yml. Otherwise, you can try creating your own environment using running the following code:

conda env create -n SyFi --file https://data.qiime2.org/distro/core/qiime2-2022.11-py38-linux-conda.yml

conda activate SyFi

mamba install -c conda-forge bioconda::salmon bioconda::spades=3.15.5 bioconda::whatshap=1.7 bioconda::bcftools=1.16 bioconda::bedtools=2.30.0 bioconda::bwa-mem2=2.2.1 bioconda::kallisto=0.48.0 bioconda::picard=2.27.5 bioconda::seqkit=2.3.1 bioconda::seqtk=1.3 bioconda::gatk

SyFi's second module (SyFi amplicon) makes use of Qiime2 to in silico extract the amplicon sequences from the SyFi-generated haplotypes, which are subsequently used to build the amplicon fingerprint. For that reason Qiime2 is first installed in the SyFi environment before installing all other remaining packages.

Executable software

For the installation of GATK, we downloaded the pre-compile software from their Github site. In the following code we use the {SOFTWARE_FOLDER_PATH} variable to define a potential software folder. Please, modify the code with your own folder path.

JAVA:

cd {SOFTWARE_FOLDER_PATH}
wget "https://download.oracle.com/java/19/latest/jdk-19_linux-x64_bin.tar.gz"
tar -xvzf jdk-19_linux-x64_bin.tar.gz
echo 'export PATH="{SOFTWARE_FOLDER_PATH}/jdk-19.0.1/bin:$PATH"' >> $HOME/.bashrc

More information on Java (jdk) here.

Download

Download the latest package release. Please, modify the code with your own folder path.

cd {SOFTWARE_FOLDER_PATH}
wget https://github.com/adriangeerre/SyFi/releases/download/v1.0/SyFi_v1.0.zip
unzip SyFi_v1.0.zip
echo 'export PATH="{SOFTWARE_FOLDER_PATH}/SyFi_v1.0/:$PATH"' >> $HOME/.bashrc

Usage

Usage: ./SyFi.sh <MODULE>

Sequential modules:
  main:      perform fingerprint identification from microbiome data.
  amplicon:  retrieve amplicon fingerprints from gene fingerprints using in silico primers.
  quant:     pseudoalign amplicon sequecing data to fingerprints.

Other:
  help:      display this help message.
  citation:  display citation.
  structure: display folder structure for execution.

Input

SyFi main assumes that the genomes and reads are organized in sub-folders inside of the input folder (-i | --input-folder). Each sub-folder should contain the genome (.fasta) and the reads (.fastq.gz).

For example:

  input_folder/
            └── strain_1
                ├── strain_1_R1.fastq.gz
                ├── strain_1_R2.fastq.gz
                └── strain_1.fasta
            └── strain_2
                ├── strain_2_R1.fastq.gz
                ├── strain_2_R2.fastq.gz
                └── strain_2.fasta
            └── ...

SyFi main loops through the samples of the folder and runs the steps in sequential order. It will run each sample one time and categorize it in Success, Skipped or Failed. Once, it runs over all samples, the option "-f | --force" must be used to re-run the sample through SyFi steps.

SyFi quant assumes that the SynCom-inoculated microbiome samples are organized in sub-folders inside of the read input folder (-i | --read-folder). Each sub-folder should contain the single or paired-end reads (.fastq.gz).

For example:

  read_folder/
            └── sample_1
                ├── sample_1_R1.fastq.gz
                └── sample_1_R2.fastq.gz
            └── sample_2
                ├── sample_2_R1.fastq.gz
                └── sample_2_R2.fastq.gz
            └── ...

Output

The default (minimum; k=0) output of SyFi consist of:

• 10-Blast/{strain}.tsv
• 11-Sequences/{strain}/{strain}.fasta
• 20-Alignment/{strain}/{strain}.fasta
• 20-Alignment/{strain}/{strain}.fastq.gz
• 30-VariantCalling/{strain}/variants/{strain}.vcf.gz
• 40-Phasing/{strain}/{strain}_assembly_h {strain}.fasta
• 40-Phasing/{strain}/{strain}_phased.vcf.gz
• 50-haplotypes/{strain}/clean_{strain}_haplotypes.fasta
• 60-Integration/{strain}/abundance.tsv
• 60-Integration/{strain}/copy_number.tsv
• 60-Integration/{strain}/integration.tsv
• 70-Fingerprints/{strain}/{strain}_all_haplotypes.fasta
• 70-Fingerprints/{strain}/seq_h {number}.fasta
• 70-Amplicon/{strain}/{strain}_all_haplotypes.fasta
• 71-Amplicon/{strain}/seq_h {number}.fasta
• 80-Pseudoalignment/{read_sample}/quant.sf
• 90-Output/copy_number.tsv
• 90-Output/raw_output_table.txt
• 90-Output/norm_output_table.txt