PATRIC Bioinformatics Resource Center Genome Assembly in PATRIC

The Sequence Assembly problem � Reconstructing contiguous DNA regions (contigs) from a set of short sequences (reads)

Sample Data: Demo FTP Server

Log on to the PATRIC website: patricbrc. org

Starting the Assembly job

Bring your own data

Launch an assembly job

Data upload and job status

Curated assembly strategies

Which assembly recipe to use � auto — the evolving default strategy recommended for � full spades - runs the full SPAdes pipeline, one of the � fast — ~2 X faster than auto; suited for large genomes or � kiki — very fast but does not use paired end � miseq — good for Illumina Mi. Seq reads that are 250 - most data best assemblers for microbial genomes simple microbial communities (velvet + megahit) information; good for metagenome assembly 350 bp long (Spades with more k-mer iterations) smart — the slowest and sometimes the most accurate � plasmid —plasmid assembly (plasmid. SPAdes) �

Typical execution times for a typical microbial genome Recipe Hours smart 3 ~ 100 auto / miseq 2 ~ 80 fast 1 ~ 12 kiki 1~6 Depends on read depth, sequencing errors, genome size, repeat structure, etc.

You can download output from the workspace

Evaluate assembled contigs

Upcoming improvements • Improved error detection and classification using supervised learning • Workflows for new sequencing technology (Min. ION, Hybrid assembly

Acknowledgements � PATRIC Team: ◦ University of Chicago ◦ FIG � Terry Disz � Ryan Aydelott � Ross Overbeek � Tom Brettin � Gordon Pusch � Neil Conrad � Veronika Vonstein � Jim Davis ◦ Stevens Group � Chris Bun � Sebastien Boisvert ◦ VBI � Emily Dietrich � Joseph Gabbard � Chris Henry � Ron Kenyon � Dan Murphy-Olson � Dustin Machi � Bob Olson � Chunhong Mao � Bruce Parrello � Bruno Sobral � Maulik Shukla � Rebecca Wattam � Rick Stevens � Andrew Warren � Fangfang Xia � Rebecca Will � Harry Yoo National Institute of Allergy and Infectious Diseases Contract No. HHSN 272201400027 C