Kneaddata

* Fix merge final file with run trf and concat pairs as prior just included orphans (used with workflow).
* Update to download new human database (hg37).

* Remove extra java option in trimmomatic call allowing for compatibility with the latest java versions.

* Relax criteria for detecting new Illumina sequence identifier to include ids that are truncated to at least 7 tokens of colon-delimited information.

* Adapter trimming has been added to the Trimmomatic step by default.
* The paired default mode of concordant plus discordant alignment has been removed and replaced with discordant alignment. An evaluation of these two modes found they are relatively the same (reads filtered differed by at most 1%) and the concordant plus discordant mode is not supported in newer versions of bowtie2. Concordant only alignments is still an option available for paired end reads.
* The default min length percent has been reduced from 70% to 50% for the Trimmomatic step.
* Newer functions (like reformat identifiers and discordant pairs) have been modified to fix error messages when running with python3.

* A new option "--remove-intermediate-output" allows for the removal of output files that were generated during the workflow. This option is helpful in reducing the total size of products generated which can be large for many samples.

* A new option "--serial" allows for alignments to be processed in a sequential manner. The clean reads from aligning to the first reference database will be used as input to the second database. The input/outputs will chain until all databases have been processed. Databases will be run in the same order as they are provided on the command line. This differs from the default mode which will run the same input files through each database, in parallel if the "--processes N" option is set to more than one.
* The log messages were updated so single end runs with multiple reference databases will have more descriptive column names, each filtered column will include the database name, when read counts are compiled from the logs with the utility script.