Let's dig further

Enhancements

• support new --isolate mode in spades 3.14
• use kmc instead of mash for faster genome size estimation
• modified some default parameters for better results
• auto-detection of cpu and ram
• overall faster performance

Bug fixes

• fix FASTA headers
• disable internal adaptor filtering in skesa
• faster contig correction
• fix ram management for some commands
• fix samtools temp file race condition clash

Into the wild we go

• Support skesa, spades, velvet and megahit as assembler engines
• Use mash instead of kmc to detect genome size
• Refine the pilon correction parameters
• Attempt to control memory usage
• Less perl dependencies
• One log file

Closer and closer...

• Major bug fix when running multiple shovills in parallel, /tmp clash removed (@andersgs)
• Remove non-core Perl module dep on File::Slurp
• TRAVIS support
• Bioconda support
• Brewsci/Bio support
• Fixed --help error codes and stdout vs stderr
• Improved docs
• Fixed typos

Let it float

• Lots of bug fixes
• New --check option to check for dependencies
• Removed non-core Perl module dep Time::Piece
• Handle different GFA names in Spades 3.11.0
• More checking of --kmers to ensure no overly long ones used
• Detect machine RAM
• Link to docker container
• Backwards compatible to older Pilon now
• FIx float/int problems with --ram and --mincov
• Increased default overlap for FLASH to 20
• Only keep .gfa file and call it contigs.gfa
• The very cool --depth subsample read feature is now default 100

Important update

• Default --depth is 100 instead of 50. The latter works good for overlapping miseq but not others.
• Delete .fastg and rename GFA to contigs.gfa
• Updated docs
• Modified log output slightly

Less is more

• Major feature - input FASTQ will be subsampled to 50x (or --depth) to reduce burden of all following steps. Increases speed a lot.
• Output fasta now has 60 characters per line
• New --namefmt to name contigs how you want it, e.g. --namefmt "EcK12.%04d.final"
• Now --help returns $? -eq 0 as it should.

RAM dammit

• New --ram option to set memory options for all the tools it can
• Better --help output
• More documentation
• Some tweaks to pilon parameters
• Trimming now default off, as most modern FASTQs are adaptor free. Use --trim if need to.

Ok, trimming works now

Forgot to push the adaptors to the repo!

Give it a go

v0.4

Add extra k value and bump to 0.4

Worth a try

v0.3

Explain trimming is default off