v0.5.0

Release Notes

• added support for Native barcodes 13-24
• upgraded to Guppy 4.4.2, medaka 1.2.0, NanoPlot 1.32.0 (important to get NanoStats.CSV file and not some weird non-standard txt file format)
• added -h option to workflow-after-gpu-basecalling.sh for displaying HELP/USAGE statement. cleaned up formatting
• run_basecalling-w-gpu.sh: now uses /scicomp/scratch instead of /tmp since it performs better
• README updates to simplify & clarify the documentation
• moved unused scripts into deprecated/ subdirs
• other minor bug fixes and tweaks to various scripts (full list of changes between v0.5.0 and v0.4.4 here)

v0.4.4

Post GIGte/v0.4.3 updates:

• switched to using singularity 2.6.1 module for all scripts
• upgraded to flye 2.7 (run via staphb container)
• cleaned up mess of guppy-logs in guppy output dir
• NanoPlot and NanoPlot --barcoded run after guppy basecalling finishes (via staphb container)
• np_consensus_racon.sh now uses minimap2 v2.17
• upgraded to medaka v1.0.1 (run via staphb container)
• updated README with new info

v0.4.3 for GIGte training

Bugs were squashed, README was cleaned up, scripts are running better.

This release is for the June 2020 GIGte training!

Curtis

v0.4.2

• Added support for:
  • R9.4.1 flowcell + rapid barcoding kit
  • R9.4.1 flowcell + ligation kit + native barcodes 1-12
  • R10 flowcell + ligation kit + native barcodes 1-12
• upgraded to Guppy v3.4.5
• qsub logs now written inside $OUTDIR/log or $OUTDIR/demux/barcode##/log for tidyness
• cleaned up README to reflect the above
  • aspen now has multiple gpu nodes with V100's
  • fixed usage statements

To do:

• Detect GPU availability since aspen GPU nodes have 2 V100 cards. guppy_basecaller -x auto is the same as saying -x cuda:0 👎
• Add support for passing in a config file to workflow-after-gpu-basecalling.sh that contains:
  • Sample ID
  • barcode number (RBK or NBD)
  • estimated genome size (to be used as input parameter in various places)
• Add support for R10.3 flowcell + ligation kit + native barcodes 1-24 (R10 flow cell discontinued)
• Allow users to specify a read length for removing reads w/ filtlong (1000 bp could be to stringent if bulk of reads are around that length)
• Test and add support for rasusa for randomly subsampling and filtering reads (filtlong is biased towards reads with highest q-scores)

v0.4.1

Fixed version of v0.4 release with correct qsub commands. Created this release, so that others can download the .tar.gz instead of git cloneing and having issues with git lfs. Repo is contstantly over it's git lfs data quota 👎

v0.4

Created this release, so that others can download the .tar.gz instead of git cloneing and having issues with git lfs. Repo is contstantly over it's git lfs data quota 👎

unit testing and standardization

• script names are more standardized
• unit testing available by editing t/environment.sh and running make test
• Salmonella Litchfield test dataset: you can assemble it and then check the assembly by serotyping it

Fire up those GPUs

This release has a script for basecalling Nanopore data on the 2 Tesla V100 GPUs available on node98, as well as a follow-up script for assembling with wtdbg2 and polishing with nanopolish via qsub

v0.1

I wanted to tag the first stable iteration of this workflow repo.