Merge pull request #105 from briney/development

Development

.travis.yml

@@ -1,11 +1,23 @@
 language: python
+
+python:
+  # - "3.6"
+  # - "3.7"
+  - "3.8"
+  - "3.9"
+
+dist:
+  - bionic
+  - focal
+
 # pick the appropriate requirements file
-matrix:
-  include:
-    - python: "3.7"
-      env: REQUIREMENTS=requirements.txt
-    - python: "3.8"
-      env: REQUIREMENTS=requirements.txt
+# matrix:
+#   include:
+#     - python: "3.8"
+#       env: REQUIREMENTS=requirements.txt
+#     - python: "3.9"
+#       env: REQUIREMENTS=requirements.txt
+
 branches:
   only:
   - master
@@ -16,7 +28,8 @@ before_install:
   - sudo apt-get install -y build-essential muscle mafft cd-hit
 # command to install dependencies
 install:
-  - pip install -r $REQUIREMENTS
+  # - pip install -r $REQUIREMENTS
+  - pip install -r requirements.txt
 # command to run tests
 script:
   - pytest # or py.test for Python versions 3.5 and below

README.md

@@ -69,7 +69,7 @@ A few helper scripts are included with abstar:
 To run the test suite, clone or download the repository and run `pytest ./` from the top-level directory.  
 
 ### requirements  
-Python 3.7+   
+Python 3.8+   
 abutils  
 biopython  
 celery  

abstar/__init__.py

@@ -1,4 +1,6 @@
-from __future__ import absolute_import, division, print_function, unicode_literals
+import warnings
+from Bio import BiopythonWarning
+warnings.simplefilter('ignore', BiopythonWarning)
 
 from .core.abstar import run, run_standalone, main, parse_arguments, validate_args
 from .preprocess import fastqc, adapter_trim, quality_trim

abstar/core/antibody.py

@@ -263,7 +263,7 @@ def _vdj_aa(self):
         # the only offset that matters is the offset between the full germline and the portion that matched.
         #self.v_rf_offset = self.oriented_input.sequence.find(self.vdj_nt) % 3
         if self.v.germline_start > 0:
-            self.v_rf_offset = 3 - self.v.germline_start
+            self.v_rf_offset = 3 - (self.v.germline_start % 3)
         else:
             self.v_rf_offset = 0
         self.coding_start = self.v.query_start + self.v_rf_offset

abstar/tests/test_indels.py

@@ -28,7 +28,7 @@ def compare_v_ins_to_reference(antibodies):
     errors = []
     for antibody in antibodies:
         ref = antibody.v
-        v = GermlineSegment(ref.full, ref.species, ref.db_name, initialize_log=False)
+        v = GermlineSegment(ref.full, ref.species, ref.db_name, 'bcr', initialize_log=False)
         v.query_alignment = ref.query_alignment
         v.germline_alignment = ref.germline_alignment
         v.alignment_midline = ref.alignment_midline
@@ -83,7 +83,7 @@ def compare_v_del_to_reference(antibodies):
     errors = []
     for antibody in antibodies:
         ref = antibody.v
-        v = GermlineSegment(ref.full, ref.species, ref.db_name, initialize_log=False)
+        v = GermlineSegment(ref.full, ref.species, ref.db_name, 'bcr', initialize_log=False)
         v.query_alignment = ref.query_alignment
         v.germline_alignment = ref.germline_alignment
         v.alignment_midline = ref.alignment_midline

abstar/tests/test_mutations.py

@@ -40,7 +40,7 @@ def compare_v_nt_mutations_to_reference(antibodies):
     for antibody in antibodies:
         ref = antibody.v.nt_mutations
         ab = MockAntibody()
-        v = GermlineSegment(antibody.v.full, antibody.v.species, antibody.v.db_name, initialize_log=False)
+        v = GermlineSegment(antibody.v.full, antibody.v.species, antibody.v.db_name, 'bcr', initialize_log=False)
         v.gene_type = antibody.v.gene_type
         v.query_alignment = antibody.v.query_alignment
         v.germline_alignment = antibody.v.germline_alignment
@@ -90,7 +90,7 @@ def compare_j_nt_mutations_to_reference(antibodies):
     for antibody in antibodies:
         ref = antibody.j.nt_mutations
         ab = MockAntibody()
-        j = GermlineSegment(antibody.j.full, antibody.j.species, antibody.j.db_name, initialize_log=False)
+        j = GermlineSegment(antibody.j.full, antibody.j.species, antibody.j.db_name, 'bcr', initialize_log=False)
         j.gene_type = antibody.j.gene_type
         j.query_alignment = antibody.j.query_alignment
         j.germline_alignment = antibody.j.germline_alignment
@@ -142,7 +142,7 @@ def compare_v_aa_mutations_to_reference(antibodies):
     for antibody in antibodies:
         ref = antibody.v.aa_mutations
         ab = MockAntibody()
-        v = GermlineSegment(antibody.v.full, antibody.v.species, antibody.v.db_name, initialize_log=False)
+        v = GermlineSegment(antibody.v.full, antibody.v.species, antibody.v.db_name, 'bcr', initialize_log=False)
         v.gene_type = antibody.v.gene_type
         v.query_alignment = antibody.v.query_alignment
         v.germline_alignment = antibody.v.germline_alignment
@@ -197,7 +197,7 @@ def compare_j_aa_mutations_to_reference(antibodies):
     for antibody in antibodies:
         ref = antibody.j.aa_mutations
         ab = MockAntibody()
-        j = GermlineSegment(antibody.j.full, antibody.j.species, antibody.j.db_name, initialize_log=False)
+        j = GermlineSegment(antibody.j.full, antibody.j.species, antibody.j.db_name, 'bcr', initialize_log=False)
         j.gene_type = antibody.j.gene_type
         j.query_alignment = antibody.j.query_alignment
         j.germline_alignment = antibody.j.germline_alignment

abstar/tests/test_regions.py

@@ -32,7 +32,7 @@ def compare_v_regions_to_reference(antibodies):
         ab = MockAntibody()
         ab.oriented_input = antibody.oriented_input
         ab.v_rf_offset = antibody.v_rf_offset
-        v = GermlineSegment(antibody.v.full, antibody.v.species, antibody.v.db_name, initialize_log=False)
+        v = GermlineSegment(antibody.v.full, antibody.v.species, antibody.v.db_name, 'bcr', initialize_log=False)
         v.query_start = antibody.v.query_start
         v.gene_type = antibody.v.gene_type
         v._imgt_position_from_raw = antibody.v._imgt_position_from_raw
@@ -65,7 +65,7 @@ def compare_j_regions_to_reference(antibodies):
         ab = MockAntibody()
         ab.oriented_input = antibody.oriented_input
         ab.v_rf_offset = antibody.v_rf_offset
-        j = GermlineSegment(antibody.j.full, antibody.j.species, antibody.j.db_name, initialize_log=False)
+        j = GermlineSegment(antibody.j.full, antibody.j.species, antibody.j.db_name, 'bcr', initialize_log=False)
         j.query_start = antibody.j.query_start
         j.gene_type = antibody.j.gene_type
         j._imgt_position_from_raw = antibody.j._imgt_position_from_raw

abstar/utils/output.py

@@ -1064,6 +1064,7 @@ def _hadoop_tabular_output(vdj):
                   'raw_input']
 
 AIRR_HEADER = ['sequence_id',
+               'umi',
                'sequence',
                'sequence_aa',
                'rev_comp',
@@ -1072,12 +1073,17 @@ def _hadoop_tabular_output(vdj):
                'stop_codon',
                'v_call',
                'v_gene',
+               'v_score',
                'v_identity',
+               'v_identity_aa',
                'd_call',
                'd_gene',
+               'd_score',
                'j_call',
                'j_gene',
+               'j_score',
                'j_identity',
+               'j_identity_aa',
                'c_call',
                'cdr3_length',
                'fwr1_aa',
@@ -1097,8 +1103,19 @@ def _hadoop_tabular_output(vdj):
                'v_deletions',
                'junction',
                'junction_aa',
+               'junction_in_frame',
+               'isotype',
+               'locus',
                'v_cigar',
                'd_cigar',
                'j_cigar',
                'species',
+               'germline_database',
                'raw_input']
+
+
+
+
+
+
+

abstar/version.py

@@ -2,4 +2,4 @@
 # 1) we don't load dependencies by storing it in __init__.py
 # 2) we can import it in setup.py for the same reason
 # 3) we can import it into your module module
-__version__ = '0.5.3'
+__version__ = '0.5.5'

requirements.txt

@@ -8,4 +8,4 @@ pandas
 pyarrow
 pymongo
 pytest
-scikit-bio
+scikit-bio==0.5.6

setup.py

@@ -59,8 +59,8 @@
                 'bin/make_basespace_credfile',],
     'include_package_data': True,
     'classifiers': ['License :: OSI Approved :: MIT License',
-                    'Programming Language :: Python :: 3.7',
                     'Programming Language :: Python :: 3.8',
+                    'Programming Language :: Python :: 3.9',
                     'Topic :: Scientific/Engineering :: Bio-Informatics']
 }