dbotu 1.4.0

Distribution-based OTU calling

dbotu3 is a new implementation of Sarah Preheim’s dbOTU algorithm. The scope is narrower, the numerical comparisons are faster, and the interface is more user-friendly.

Read the documentation for:

• a guide to getting started,

• an explanation of the algorithm, and

• the API reference.

You can also read our new paper for more technical details about the algorithm. The Alm Lab website also has a short page with information.

Installation

dbotu3 is on PyPi and can be installed with pip.

Requirements

• Numpy, SciPy, BioPython, Pandas

• Levenshtein

Version history

• 1.1: Corrected error where sequence IDs that could be read as integers would not be found in the table

• 1.2: Python 2 compatibility, tox test framework, warnings for improperly-formatted sequence count tables

• 1.2.1: Added setup requirements

• 1.3.0: Improved OTU file header. Split the log file into a debug and progress log.

• 1.4.0: Made an improvement to the Levenshtein-based genetic dissimilarity metric.

To-do

• Fix the output format (maybe put OTU_ID in the first field and row?)

• Find a way to avoid reading the entire count file? (The fasta file is not all loaded because we use SeqIO.index.)

• Better coverage for unit tests

Author

Scott Olesen / swo at alum.mit.edu