Supernova on Flashlite @ QCIF / RCC

Accessing tool/workflow

This repository contains PBS scripts that run the globally installed supernova on University of Queensland's HPC, Flashlite. Supernova (v2.1.1) has been installed as a system module and can be accessed with the command:

module load supernova/2.1.1

Quickstart tutorial

The supernova.pbs runs:

• supernova run separately for each sample to generate a whole genome de novo assembly for each.
• supernova mkoutput in order to generate various styles of FASTA output for your assemblies including:
  • --style=raw
  • --style=megabubbles
  • --style=pseudohap
  • --style=pseudohap2

10X data

1. Clone this repository into your working directory on Flashlite by:

git clone https://github.com/Sydney-Informatics-Hub/Flashlite-Supernova.git

2. In Flashlite-Supernova, edit the variables in supernova.pbs including the:

• FQ_PATH
• RUN_ID
• OUTDIR

3. Run the script by qsub supernova.pbs

TELLSeq data

Supernova can be used to assemble 'Transposase Enzyme Linked Long-read Sequencing' (TELLseq) data. You’ll need to install a local version of supernova on /30days or /90days, and then to replace the 10X barcode file with the TELLSeq provided barcode file before running the scripts. Refer to TELL-Seq Data Analysis Roadmap User Guide.

Optimisation required

The scripts utilise local to the node disk $TMPDIR which handles I/O better than other filesystems.

Infrastructure usage and benchmarking

Summary

Exemplar 1: e.g. Cockroach (TELLSeq data)

Species	CPUs_requested	CPUs_used	CPU_percent	Mem_requested	Mem_used	VMem_used	CPUtime	CPUtime_mins	Walltime_req	Walltime_used	Walltime_mins	JobFS_req	JobFS_used	Efficiency	Service_units(1*CPU_hours)	Queue	Account	ExitStatus
Cockroach	24	24	2400	496gb	158606368kb	305991020kb	830:26:04	49826.07	336:00:00	61:34:24	3694.4	NA	NA	0.56	830.43	Long	qris-user	0

Cite us to support us!

The Flashlite-Supernova pipeline can be cited as DOI: https://doi.org/10.48546/workflowhub.workflow.151.1