v0.12.1

RELEASE NOTES FOR FastQC v0.12.1

This is a minor bugfix release. It fixes a problem with gzip
file detection on OSX where the java probeContentType can't
tell that a gzipped file isn't just a binary stream.

v0.12.0

RELEASE NOTES FOR FastQC v0.12.0

This release introduces new functionality as well as resolving some
existing problems.

• Fix bugs in the per tile plot if zero length sequences are present

• Add a count of total bases to the Basic Stats output

• Use better sorting of the best contaminant finding

• Add a "dup_length" option to specify the length of sequence used
  for detecting duplicates

• Made the default duplicate detection length 50bp, regardless of
  the length of the library. Previously only sequences over 75bp
  were truncated, now the limit applies to everything

• Removed the plot of deduplicated duplication levels from the
  duplication plot since it just confused people and everyone ignored
  it.

• Fixed a bug in iterating through fast5 files in multiple directories

• Fixed a documentation bug in the Duplicate plot

• Improved memory handling - default allocation is now 512MB and there
  is a --memory option which can increase the allocated memory without
  having to play with --threads

• Colours in the per base sequence composition plot have been changed
  to be more colourblind friendly

• BAM parsing now uses htsjdk on the back end

• Add a --svg option to make the report use SVG graphics instead of
  PNG. The zip file will always contain both PNG and SVG versions of
  all figures

• Add line numbers to error messages where parsing errors occur

• Add an option to delete the zip file if it is uncompressed during
  processing (add --delete on top of --extract)

• The default adapters removed the SOLID adapter, but added polyA
  and polyG since these can both provide useful information and are
  often trimmed from sequences in a similar manner to adapters.

v0.11.9

This is a bugfix release which resolves some issues with the program;

• We removed the native look and feel from the linux application since
  it's horribly broken

• Fixed a hang if a run terminated from an out-of-memory error

• Bundle a suitable JRE with the OSX app build

• Fixed a corner case where adapters could occasionally be double-counted

• Updated the fast5 parser to account for the newer format multi-read
  oxford nanopore fast5 files

• Fixed problems if analysing a completely blank file

v0.11.8

This release works around some edge cases in unusual sequence libraries
and changes the behaviour of the read length module when run with the
--nogroup option. Other minor fixes are also present.

v0.11.7

This is a bugfix release for a bug introduced in 0.11.6. Specifically
this version would crash if the first sequence in a file was <12bp
(or less than the length of the longest adapter if a custom adapters
file was being used).

v0.11.6

RELEASE NOTES FOR FastQC v0.11.6

This update fixes some bugs and updates some of the functionality to
accommodate changes in some of the sequencing platforms.

There is one major change which is that by default we now disable the
kmer module. With the inclusion of the adapter plot the value of the
information in the Kmer plot is often not great, and it is easy to
confound it if there are any over-represented sequences, or primer
compositional bias. Overall therefore we consider it best to not
routinely include this module.

If you want to turn this module back on, then simply edit the
limits.txt file in the Configuration folder of the FastQC installation
and change the line near the top which says:

kmer ignore 1

..to..

kmer ignore 0

..and the module will be re-enabled.

Other changes in this release are:

1. Fixed a bug which prematurely abandoned the adapter content plot when long custom adapters were being used.

2. Changed the cutoff for the maximum number of tiles to allow for the novaseq which has lots of them.

3. Fixed a bug in the parsing of tile numbers on some illumina sequencers

4. Added some new Clontech sequences to the contaminants list.

5. Made the --nanopore option work with the new multi-folder ONT folder structure

6. Added an option to specify a file name when streaming data into FastQC

7. Added new RDF paths to check for fastq data in nanopore fast5 files

8. Fix parsing of newer nanopore base names to correctly collate sequences

9. Fixed a typo in the documentation for the per tile plot documentation

10. Added a --min-length option to ignore short sequences making it easier to generate directly comparable statistics between runs.