ProSynTax: Prochlorococcus and Synechococcus Taxonomy Dataset

INTRODUCTION:

Understanding the distribution of the abundant and closely related picocyanobacteria, Prochlorococcus and Synechococcus, is essential for understanding marine ecosystems. These organisms are highly diverse, making the accurate classification of clusters/clades/grades within each genus challenging. As a result, Prochlorococcus and Synechococcus populations are often characterized as a single strain, concealing the well-documented fine-scale niche partitioning within these groups (Johnson, et al., 2006, Hunter-Cevera, et al., 2016, Larkin, et al., 2016, Kent, et al., 2019, Thompson, et al., 2021, Ustick et al., 2023). Here, we introduce ProSynTax and its associated workflow, designed to significantly improve metagenomic classification of Prochlorococcus and Synechococcus using a substantial amount of high-quality genomic reference data collected over the past decade and from this study. ProSynTax includes 1,260 single-cell amplified genomes, high-quality draft cultured genomes, and unpublished closed genomes, featuring new closed circular assemblies for 39 Prochlorococcus, 12 Synechococcus, and 12 marine heterotrophic bacterial strains. This includes 29 Prochlorococcus genomes that were previously partially assembled (Biller et al., 2014) and 10 genomes from unpublished isolates. Additionally, the dataset includes 41,753 genomes of marine heterotrophic bacteria, archaea, and viruses to assess communities surrounding Prochlorococcus and Synechococcus. ProSynTax and the accompanying workflow can accurately identify clades in metagenomic samples containing at least 0.15% Prochlorococcus reads or 0.03% Synechococcus reads, thereby improving our understanding of these picocyanobacteria in low-abundance regions.

Github to the associated workflow: https://github.com/jamesm224/ProSynTax-workflow

FILE DESCRIPTION: 

Reference genomes in ProSynTax: 

ProSynTax_genomes.csv Table of genomes included in the ProSynTax dataset and their associated metadata. Data fields are as follows: organism: The name of the organism recorded in NCBI when available. For genomes/organisms obtained from sources other than NCBI, the organism’s name is provided in NCBI format genome_short_name: The genome name used in the ProSynTax datasetdomain: Bacteria, Archaea, Eukarya, or Virus genus: The genus of the organism in NCBI clade: The major cluster/clade/grade of Prochlorococcus or Synechococcus based on phylogenetic reconstruction using a concatenated alignment of proteins encoded by single-copy core genes NCBI_BioProject: The NCBI BioProject accession number associated with the organism, when available NCBI_BioSample: The NCBI BioSample accession number associated with the organism, when available NCBI_GenBank: The NCBI GenBank accession number associated with the genome sequence data, when available IMG_Genome_ID: The IMG Genome ID accession number, also known as the IMG Taxon ID, corresponds to the genome or organism in the Joint Genome Institute’s (JGI) Integrated Microbial Genomes (IMG) repository, when available New_genome_SRA_accession: SRA accession numbers for new genomes generated for this study 

Taxonomic classification files for Kaiju: 

ProSynTax_names.dmp Names taxonomy file for use with the ProSynTax dataset

ProSynTax_nodes.dmp Nodes taxonomy file for use with the ProSynTax dataset

ProSynTax_v1.1.fmi.bz2 Index file containing contents of ProSynTax_v1.faa for use with ProSynTax dataset 

Raw protein files used to build .fmi file for Kaiju: 

ProSynTax_v1.1.faa.bz2 File containing protein sequences used by Kaiju for classification of reads. Each protein sequence contains a header starting with “>” 

ProSynTax_v1.1_without_refseq.faa.bz2 File containing protein sequences found in ProSynTax with NCBI RefSeq genomes removed. Each protein sequence contains a header starting with “>”. 

Files for read count normalization: 

CyCOG6.dmnd Database containing orthologous groups of proteins used in the cluster/clade/grade normalization step  

average_cycog_length.csv Comma separated file containing the average length for each protein sequence used in the normalization step. Data fields are as follows:  cycog: name for single-copy core gene mean_AA_length: the average length of amino acids in the protein sequence of the gene 

Technical validation files: 

ProSynTax-workflow_benchmarking_genomes.tsv This tab-delimited file contains a list of subsetted genomes used in each benchmarking experiment reported in the Technical Validation section. Data fields are as follows:Experiment Name: name of benchmarking experiment conducted Subset ID: unique ID from the random genome subsetting  Genome Name: name of genome used in benchmarking experiment  

ProSynTax-workflow_benchmarking_composition.tsv This tab-delimited file contains the taxon composition of all samples used in each benchmarking experiment reported in the Technical Validation section. Data fields are as follows: Experiment Name: name of benchmarking experiment conducted Sample Name: unique sample name  Percent Prochlorococcus: percent of reads in simulated sample originating from Prochlorococcus genomes  Percent Synechococcus: percent of reads in simulated sample originating from Synechococcus genomes  Percent Heterotroph: percent of reads in simulated sample originating from marine heterotrophic bacterial genomes  Experiment Description: description of each benchmarking experiment which corresponds to varying taxonomic compositions Notes: additional information about the simulated sample