SPAdes assembler

This preprint is a draft. But given the immense utility of Logan for biology and medicine, we are immediately/freely releasing the data for early-adoption. Reach out if you’d like to collaborate. #Genomics is our shared evolutionary heritage, and so we must keep it free.

While analyzing millions of gigabytes of data may seem insurmountable today, at one point so was analyzing an entire human genome. Open source #bioinformatics illuminated the #genomic era, and now we invite the community to enter the #petabase era together.

To accomplish petabase-scale sequence assembly, we deployed a massive #AWS Batch #cloudcomputing workflow, reaching peak simultaneous-usage of 2.18 million vCPUs. After a year of preparation, the runs were executed in only 30 hours.

Logan is the assembly of all 27M DNA/RNA-seq data (44+ petabases). It expands the available genetic diversity across the Tree of Life by orders of magnitude. And it enabled us to perform a true pan-SRA sequence search in only 11 hours. Data link (cc0): registry.opendata.aws/pasteu…

Today we’re excited to freely share an early-version of, perhaps, the world’s most expansive genetics dataset: Logan. #bioinformatics #petabase #genetics #genomics #openscience biorxiv.org/content/10.1101/…

Serratus is published in @Nature!🎉 Planetary viral discovery by analysis of all available RNA sequencing data. nature.com/articles/s41586-0…

Preprint: biorxiv.org/content/10.1101/…

Special thanks to @awscloud and @UBC #CIC for project support and the SRA Team @NIHDataScience! <3

This thread is on behalf of: Artem Babaian, @robertedgarphd, Jeff Taylor, @tomeraltman, @akorobeynikov, @meleshko_da, @PierreBarbera, @themicrobeguy, @banfieldlab, @NovakovskyG, @victorlin_ , @danlohrdev, @bbuchfink, and @Marcos_dlP

Serratus is a collaborative #OpenScience project started at the @hackseq hackathon. If you’re a scientist/dev interested in doing cool research, join our Slack! All hands welcome. join.slack.com/t/hackseq-rna…

Serratus is not only optimized for analyzing the data available today, but also to keep up with the exponential growth of sequencing for tomorrow.

@bbuchfink, author of the amazing DIAMOND software for translated-nucleotide search (github.com/bbuchfink/diamond), specifically tweaked it for the Serratus use-case, dropping CPU-runtime by >10x.

@Marcos_dlP tirelessly recovered a cool 54 novel Deltavirus / Delta-like viruses. Then stumbled on 300+ truly enigmatic “Zeta Viruses”, blurring the lines of the early evolutionary origins of deltaviruses.

With a state-of-the-art assembler, we identify a clade of corona-like viruses related to PsNV (shoutout @gidmord) and show that these viruses are likely encoded on segmented genomes.

@akorobeynikov and @meleshko_da pushed the limits de novo assembly with 'coronaSPAdes', allowing for synteny-informed graph traversing. This greatly improves virus assembly, learn more in the lovely companion piece: biorxiv.org/content/10.1101/…

In a companion manuscript, the Palmscan algorithm is explored for extracting palmprints: biorxiv.org/content/10.1101/….

To distinguish so many viruses we developed a structurally informed “barcode sequence” encompassing the catalytic core of RdRP we call the “Palmprint”. It’s like 16S sequencing, but for RNA viruses.

In total we assembled aligned-hits from 3M libraries to yield 800,000 RdRP contigs which clustered into 130,000 novel species-like OTU (90% amino acid identity).

The entire search was completed in 11-days wall-clock. All software is open-source and the data is freely available. Explore the Open Virome at serratus.io