Approximately 3.9 billion contigs were created to support the Virus Discovery Hackathon, coordinated by NCBI and hosted by SDSU on Jan 9, 2019 1. SRA data selection is described in the associated publication, but, briefly, runs were selected to enrich for WGS metagenomic studies likely to contain crassphage. Contigs were then assembled using SKESA 2. First, human reads were filtered out via HISAT2 3. Next, viral contigs were assembled via guided assembly. Finally, the remaining reads were assembled, if possible, via de novo assembly. How to access these contigs and the naming conventions used are described below. Additionally, descriptive metadata, of the contigs, and the associated biosample records was put in tabular format, with details on how to access this information included below as well. Currently the information is hosted in both Amazon and Google’s cloud environments, as reflected in the access instructions provided here.
If publishing results which make use of these contigs please cite: Connor R, Brister R, Buchmann JP, et al. NCBI's Virus Discovery Hackathon: Engaging Research Communities to Identify Cloud Infrastructure Requirements. Genes (Basel). 2019;10(9)
All the contigs derived from a single SRR are included in a single fasta file. The naming convention for these files is: SRR_ACCESSSION.fa
Within each fasta the definition line of each contig is constructed as follows: SRR_ACCESSION.CONTIG_NAME.NUMBER Where CONTIG_NAME can take one of two formats. For guided assemblies RefSeqID_NUMBER, and for de novo assemblies Contig_NUMBER_NUMBER.NUMBER
The contig sequences are available from GCP and AWS platforms. In general it is preferable to work from a machine in the same region the data is located.
The bucket containing the data is experimental-sra-metagenome-contigs and it is located in Multi-region.
Examples of how the data might be accessed:
gsutil cp -r gs://experimental-sra-metagenome-contigs/ <dest>
wget https://storage.googleapis.com/experimental-sra-metagenome-contigs/DRR000019.contigs.fasta
The bucket containing the data is experimental-sra-metagenome-contigs and is located in us-east-1.
Examples of how the data might be accessed:
aws s3 cp s3://experimental-sra-metagenome-contigs.s3.amazonaws.com/DRR000019.contigs.fasta <dest>
wget http://experimental-sra-metagenome-contigs.s3.amazonaws.com/
Metadata associated with the contigs can be found in both GCP and AWS environments. The available metadata fields are outlined below. For blast related statistics, blastn was used and only viral targets were checked.
| field | description |
|---|---|
| accession | the accession for the SRR the contig is derived from |
| contig | the contig ID |
| defline | the defline seen in the associated fasta file |
| type | either guided_contig or denovo_contig |
| guide | the accession for the guide sequence if type is guided_contig |
| mean_coverage | average coverage for the contig |
| contig_length | contig length in bases |
| subject | accession for best blast hit |
| subject_taxid | taxid for best blast hit |
| subject_title | defline for best blast hit |
| pident | percent identity for best blast hit |
| evalue | evalue for best blast hit |
| bitscore | bitscore for best blast hit |
| alignment_length | alignment length for best blast hit |
The metadata is loaded into BigQuery and can be accessed via the web console by pinning the project research-sra-cloud-pipeline and opening the dataset realign and looking at the table experimental_sra_metagenome_contigs
The metadata is loaded into s3 and can be accessed via the Athena web console by creating a new database. First select query data in Amazon S3, then Add a table manually. Next, specify the location of the input data set as s3://experimental-sra-metagenome-contigs-us-east-1/metadata/. Finally, select csv as the data format followed by selecting bulk add columns and pasting the field below into the text box.
accession string,
contig string,
defline string,
type string,
guide string,
mean_coverage float,
contig_length int,
subject string,
subject_taxid int,
subject_title string,
pident float,
evalue float,
bitscore float,
alignment_length int