Need to identify and analyze each organism in a sample (isolate or polymicrobial)? We built a highly-accurate method to tease apart the organisms in a #metagenomic assembly, enabling organism-specific AMR, abundance estimation, strain-typing & more: doi.org/hhqr 🧵1/7
We show that aligning contigs against a nucleotide reference database enables far more precise classification of assembled sequences, surpassing amino acid alignments. 2/7 
For example, on an isolate of Bacillus anthracis, BugSplit identified the organism to species, whereas protein-based tools (MMSeqs2, DIAMOND+MEGAN-LR) identified the organism to genus (Bacillus). 3/7
Applied to @nanopore and @PacBio #metagenomic data, BugSplit produces leading accuracy estimates on the presence and abundance of organisms in mixed samples. External benchmarks already support these findings (credit to an amazing @PacBio team and co): doi.org/hhqs 4/7
In collaboration with researchers at @FHIOxford @MangesLab @VCHhealthcare @CDCofBC, we sequenced respiratory samples containing SARS-CoV-2 and show that BugSplit can detect a novel coronavirus with a reference database predating the pandemic (ie. not containing SARS-CoV-2). 5/7
We applied BugSplit to @nanopore #metagenomic sequencing of urine and demonstrate that BugSplit can accurately predict antimicrobial resistance of Neisseria gonorrhoeae directly from sample (ie. no culture!). More examples (eg. blood cultures) in the paper. 6/7
Lastly, we want to enable BugSplit for all. BugSplit is set up, cloud-accelerated and ready to use at bugseq.com. 7/7
• • •
Missing some Tweet in this thread? You can try to
force a refresh
