Efficient sequence alignment path data structure (scikit-bio#2011)

* initial commit

* add path function to tabularMSA and create util file

* Add _util.py file

* Add basic to_cigar and from_cigar functions

* Start unit tests for PairAlignPath

* Add ability to handle = or X to from_cigar function

* Add code attribution for part taken from SO

* Remove print statement

* Initial version of handling match vs mismatch for to_cigar

* Split encoding into separate function

* Change input name

* Add test data

* Start on unit tests

* Update fix_arrays function

* Change from np.nan to 0 for append in fix_arrays

* Numpy version of run_length_encode

* Fix fix_arrays function

* Enable from_cigar to handle strings with or without ones

* Add error handling and tests for from_bits in PairAlign

* PairAlignPath fully covered

* Expand unit tests

* Test more than 8 seqs for from_bits

* To_indices tests

* Complete coverage for to_indices

* Full coverage

* Update init file

* Added non default gap character handling to from_tabular

* Add skbioobject to class

* Update fix_arrays

* Basic repr function

* Switch to starts from n_seqs

* Rewrite unit tests

* Remove large data files

* Simplify pairwise_align_score

* Docstring as raw strings and fix toctree

* Simplify pairwise align_score and rename multiple_align_scores

* Fix tabular from_path_seqs

* Improve docstrings

* Change to np.int64 to make Windows compatible

* Remove alignment score functions for future PR

* Allow for non numpy array like in from_bits

* Remove overrides decorator, see what happens

* Finish docstrings for PairAlignPath

* Remove unused import

* Improve error handling for to_cigar and remove subset

* Improve error handling for to_cigar

* Figuring out to_cigar

* Fix to_cigar functionality by converting to str

* Complete unit tests for to_cigar

* Finish unit test for from_cigar

* Unit tests for initialization

* Modify Pair to_bits to handle two gaps

* Functional version of to_cigar

* Enhanced to_cigar function

* Remove gap chars for to_cigar

* Enhancement comments

* Modify to_cigar logic

* Optimized version of to_cigar

* Update docstrings

* Update unittests

* Test RLE

* Add Tabular from_path_seqs test

* Remove float option for gap in to_indices

* Update CHANGELOG

* Address most recent comments

* Move mapping and switch to unsigned int for starts

* Rename mapping and codes

* Create class properties for states, starts, lengths, and shapes

* Switch ValueError to TypeError where appropriate

* Paired programming additions

* Update to/from_indices functionality to handle starts

* Update to/from_coordinates functionality to handle starts

* Lint tests

* Remove unused import

* Start on docstring examples

* AlignPath docstring

* Add examples to docstrings

* Add example text

* More examples

* Final examples

* Change gaps to starts

* Fix doctests

CHANGELOG.md

@@ -5,6 +5,7 @@
 ### Features
 
 * Added `ProteinEmbedding` class and corresponding file format ([#2008](https://github.com/scikit-bio/scikit-bio/pull/2008]))
+* Added `AlignPath` and `PairAlignPath` classes ([#2011](https://github.com/scikit-bio/scikit-bio/pull/2011))
 
 ## Version 0.6.0
 

skbio/alignment/__init__.py

@@ -15,6 +15,8 @@
    :toctree: generated/
 
    TabularMSA
+   AlignPath
+   PairAlignPath
 
 
 Alignment algorithms
@@ -218,11 +220,14 @@
     local_pairwise_align_ssw,
 )
 from skbio.alignment._ssw_wrapper import StripedSmithWaterman, AlignmentStructure
+from skbio.alignment._path import AlignPath, PairAlignPath
 
 __all__ = [
     "TabularMSA",
     "StripedSmithWaterman",
     "AlignmentStructure",
+    "AlignPath",
+    "PairAlignPath",
     "local_pairwise_align_ssw",
     "global_pairwise_align",
     "global_pairwise_align_nucleotide",