skmer 3.3.0

• Now libraries created with skmer 3.0.2 or older whose random seed was not explicitly set, can be reused and mixed with newer libraries which their seed is 42

skmer 3.2.1

• Adding two new commands to compute branch supports for phylogenies generated from estimated distances
  • subsample creates a subsample folder containing replicate subfolders
  • correct performs distance correction on subsampled replicates

skmer 3.1.0

• Increasing the minimum required version of Mash to v2.3. Apparently they have fixed something showing improved accuracy in our tests.
• Adding -S option to reference command to allow the user specify the random seed used for sketching. Libraries of different seeds can't be merged. Default value is 42, which is consistent with older sketches.
• Changing the default sketch size to 100000

skmer 3.0.2

• Fixing an issue with -a option in query sub-command, when the library and working directory are on different file systems (#18 )

skmer 3.0.1

• Fixing an internal bug in reading estimated parameters

• Suppressing the exception raised when the coverage is too low to estimate

skmer 3.0.0

• Detecting the type of input file (assembled sequences or unassembled reads) from the length of sequences.

• Switching off the correction if the coverage is found to be very low and hard to be estimated robustly.

• Outputting distance 5.0 whenever JC transfromation is on and the estimated distance is above 0.75

skmer 2.0.2

• Updating the displayed version and input format in the help message

• Resolving a bug in the error message when the distance is out of range for JC transformation

skmer 2.0.1

• Increasing the minimum required version of numpy to suppress a warning (Currently, installation on Linux through anaconda installs numpy 1.15.0 which throws the warning, apparently it can be safely ignored but you can also manually upgrade numpy by running condo install numpy=1.15.1).

• Also increased the minimum required version for scipy and pandas.

skmer 2.0.0

• Accepting FASTA files as input (#5)

• Supporting all standard file extensions for FASTA/Q files (#10)

• Adding distance sub-command to compute distances for a processed library (#9)

• Improving error handling and error messages (#7)

• Fixing the bug with keyboard interrupts (#14)

• Fixing the bug with Jukes-Cantor at d=0 (#11)

• Using the average when the read length is not constant (#13)

skmer 1.1.0

• Fixing bugs with python 3
• Improving the resource usage in multi-processing mode