Merge pull request #38 from nleroy917/hotfix-sdk

Fix Import Issues For PyPi Package

.github/workflows/black.yml

@@ -0,0 +1,11 @@
+name: Lint
+
+on: [pull_request]
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - uses: actions/setup-python@v2
+      - uses: psf/black@stable

main.py

@@ -17,11 +17,12 @@
     allow_headers=["*"],
 )
 
+
 @app.get("/")
 async def root():
     return {
-        'message': "Optipyzer codon optimization server. For more details see https://github.com/nleroy917/optipyzer",
-        'url': 'https://github.com/nleroy917/optipyzer',
-        'version': VERSION,
-        'fastapi_version': fastapi.__version__
-    }
+        "message": "Optipyzer codon optimization server. For more details see https://github.com/nleroy917/optipyzer",
+        "url": "https://github.com/nleroy917/optipyzer",
+        "version": VERSION,
+        "fastapi_version": fastapi.__version__,
+    }

optipyzer/const.py

@@ -26,44 +26,92 @@
 }
 
 AMINO_ACID_LOOKUP = {
-    'ATA':'I', 'ATC':'I', 'ATT':'I', 'ATG':'M',
-    'ACA':'T', 'ACC':'T', 'ACG':'T', 'ACT':'T',
-    'AAC':'N', 'AAT':'N', 'AAA':'K', 'AAG':'K',
-    'AGC':'S', 'AGT':'S', 'AGA':'R', 'AGG':'R',                
-    'CTA':'L', 'CTC':'L', 'CTG':'L', 'CTT':'L',
-    'CCA':'P', 'CCC':'P', 'CCG':'P', 'CCT':'P',
-    'CAC':'H', 'CAT':'H', 'CAA':'Q', 'CAG':'Q',
-    'CGA':'R', 'CGC':'R', 'CGG':'R', 'CGT':'R',
-    'GTA':'V', 'GTC':'V', 'GTG':'V', 'GTT':'V',
-    'GCA':'A', 'GCC':'A', 'GCG':'A', 'GCT':'A',
-    'GAC':'D', 'GAT':'D', 'GAA':'E', 'GAG':'E',
-    'GGA':'G', 'GGC':'G', 'GGG':'G', 'GGT':'G',
-    'TCA':'S', 'TCC':'S', 'TCG':'S', 'TCT':'S',
-    'TTC':'F', 'TTT':'F', 'TTA':'L', 'TTG':'L',
-    'TAC':'Y', 'TAT':'Y', 'TAA':'_', 'TAG':'_',
-    'TGC':'C', 'TGT':'C', 'TGA':'_', 'TGG':'W',
+    "ATA": "I",
+    "ATC": "I",
+    "ATT": "I",
+    "ATG": "M",
+    "ACA": "T",
+    "ACC": "T",
+    "ACG": "T",
+    "ACT": "T",
+    "AAC": "N",
+    "AAT": "N",
+    "AAA": "K",
+    "AAG": "K",
+    "AGC": "S",
+    "AGT": "S",
+    "AGA": "R",
+    "AGG": "R",
+    "CTA": "L",
+    "CTC": "L",
+    "CTG": "L",
+    "CTT": "L",
+    "CCA": "P",
+    "CCC": "P",
+    "CCG": "P",
+    "CCT": "P",
+    "CAC": "H",
+    "CAT": "H",
+    "CAA": "Q",
+    "CAG": "Q",
+    "CGA": "R",
+    "CGC": "R",
+    "CGG": "R",
+    "CGT": "R",
+    "GTA": "V",
+    "GTC": "V",
+    "GTG": "V",
+    "GTT": "V",
+    "GCA": "A",
+    "GCC": "A",
+    "GCG": "A",
+    "GCT": "A",
+    "GAC": "D",
+    "GAT": "D",
+    "GAA": "E",
+    "GAG": "E",
+    "GGA": "G",
+    "GGC": "G",
+    "GGG": "G",
+    "GGT": "G",
+    "TCA": "S",
+    "TCC": "S",
+    "TCG": "S",
+    "TCT": "S",
+    "TTC": "F",
+    "TTT": "F",
+    "TTA": "L",
+    "TTG": "L",
+    "TAC": "Y",
+    "TAT": "Y",
+    "TAA": "_",
+    "TAG": "_",
+    "TGC": "C",
+    "TGT": "C",
+    "TGA": "_",
+    "TGG": "W",
 }
 
 AA_CODON_LIBRARY = {
-    'A': ['GCT','GCC','GCA','GCG'],
-    'R': ['CGT','CGC','CGA','CGG','AGA','AGG'],
-    'N': ['AAT','AAC'],
-    'D': ['GAT','GAC'],
-    'C': ['TGT','TGC'],
-    'Q': ['CAA','CAG'],
-    'E': ['GAA','GAG'],
-    'G': ['GGT','GGC','GGA','GGG'],
-    'H': ['CAT','CAC'],
-    'I': ['ATT','ATC','ATA'],
-    'L': ['TTA','TTG','CTT','CTC','CTA','CTG'],
-    'K': ['AAA','AAG'],
-    'M': ['ATG'],
-    'F': ['TTT','TTC'],
-    'P': ['CCT','CCC','CCA','CCG'],
-    'S': ['TCT','TCC','TCA','TCG','AGT','AGC'],
-    'T': ['ACT','ACC','ACA','ACG'],
-    'W': ['TGG'],
-    'Y': ['TAT','TAC'],
-    'V': ['GTT','GTC','GTA','GTG'],
-    'Stop': ['TAA','TAG','TGA']
-}
+    "A": ["GCT", "GCC", "GCA", "GCG"],
+    "R": ["CGT", "CGC", "CGA", "CGG", "AGA", "AGG"],
+    "N": ["AAT", "AAC"],
+    "D": ["GAT", "GAC"],
+    "C": ["TGT", "TGC"],
+    "Q": ["CAA", "CAG"],
+    "E": ["GAA", "GAG"],
+    "G": ["GGT", "GGC", "GGA", "GGG"],
+    "H": ["CAT", "CAC"],
+    "I": ["ATT", "ATC", "ATA"],
+    "L": ["TTA", "TTG", "CTT", "CTC", "CTA", "CTG"],
+    "K": ["AAA", "AAG"],
+    "M": ["ATG"],
+    "F": ["TTT", "TTC"],
+    "P": ["CCT", "CCC", "CCA", "CCG"],
+    "S": ["TCT", "TCC", "TCA", "TCG", "AGT", "AGC"],
+    "T": ["ACT", "ACC", "ACA", "ACG"],
+    "W": ["TGG"],
+    "Y": ["TAT", "TAC"],
+    "V": ["GTT", "GTC", "GTA", "GTG"],
+    "Stop": ["TAA", "TAG", "TGA"],
+}

optipyzer/db/database.py

@@ -9,4 +9,4 @@
 )
 SessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=engine)
 
-Base = declarative_base()
+Base = declarative_base()

optipyzer/db/interfaces.py

@@ -12,16 +12,17 @@ def _get_db():
     finally:
         db.close()
 
+
 def get_autocomplete_organisms():
     """Return list of organisms for the autocomplete component"""
-    
+
     with SessionLocal() as db:
         return db.query(AutocompleteOrganism).all()
-    
+
 
 def calc_codon_usage(organism_id: int):
     """
-    Function to calculate the codon usage data for a given organism. 
+    Function to calculate the codon usage data for a given organism.
     The codon usage data in this context is the percentage that a particular
     codon is used when coding for a particular amino acid.
 
@@ -48,7 +49,11 @@ def calc_codon_usage(organism_id: int):
 
         # Check for data issues (Sometings codon usage data is low, and must be dealt with)
         if sum_codons == 0:
-            print('Not enough codon data found... Skipping Organism {}'.format(organism_id))
+            print(
+                "Not enough codon data found... Skipping Organism {}".format(
+                    organism_id
+                )
+            )
             break
 
         # define index or and new sub-dictionary
@@ -69,7 +74,7 @@ def calc_codon_usage(organism_id: int):
         for codon in codon_usage[aa]:
             codon_usage[aa][codon] /= sum_codons
 
-    return counts, codon_usage # Return this dictionary of dictiomaries
+    return counts, codon_usage  # Return this dictionary of dictiomaries
 
 
 def amino_acid_usage(org_id: int, aa: str):
@@ -87,12 +92,14 @@ def amino_acid_usage(org_id: int, aa: str):
         cols = [getattr(CodonUsage, c) for c in codons]
         return db.query(*cols).filter(CodonUsage.org_id == org_id).first()
 
+
 def search_for_species(name: str) -> list[Organism]:
     """Search for an organism based on a name"""
     with SessionLocal() as db:
         return db.query(Organism).filter(Organism.species.ilike(f"%{name}%")).all()
 
+
 def get_species_by_id(org_id: str) -> Organism:
     """return an organism object given an org_id"""
     with SessionLocal() as db:
-        return db.query(Organism).get(org_id)
+        return db.query(Organism).get(org_id)

optipyzer/db/models.py

@@ -4,6 +4,7 @@
 
 from sqlalchemy import Column, Integer, String, Float
 
+
 class Organism(Base):
     __tablename__ = "organisms"
 
@@ -21,11 +22,12 @@ class Organism(Base):
     GC2_perc = Column(Float)
     GC3_perc = Column(Float)
 
+
 # this is a truncated version of the above table
 # to improve browser performance
 class AutocompleteOrganism(Base):
     __tablename__ = "autocomplete_data"
-    
+
     org_id = Column(Integer, primary_key=True)
     division = Column(String)
     assembly = Column(String)
@@ -41,8 +43,9 @@ class AutocompleteOrganism(Base):
     GC3_perc = Column(Float)
     category = Column(String)
 
+
 class CodonUsage(Base):
-    
+
     __tablename__ = "codon_usage"
 
     org_id = Column(Integer, primary_key=True)

optipyzer/dependencies.py

@@ -5,6 +5,7 @@
 from .utils import clean_seq
 from .request_models import OptimizeQuery
 
+
 def verify_dna(query: OptimizeQuery) -> OptimizeQuery:
     """
     Verify that a DNA sequence complies with the following:
@@ -19,14 +20,15 @@ def verify_dna(query: OptimizeQuery) -> OptimizeQuery:
     # is divisible by 3 (codons)
     if len(query.seq) % 3 != 0:
         raise HTTPException(400, detail="DNA sequence must be devisible by 3")
-    
+
     # verify all characters are
     # ATGC
     for base, i in zip(query.seq, range(len(query.seq))):
         if base.upper() not in "ATGC":
             raise HTTPException(400, f"Invalid base '{base}' in query at position: {i}")
     return query
 
+
 def verify_protein(query: OptimizeQuery) -> OptimizeQuery:
     """
     Verify that a protein sequence complies with the following:
@@ -41,14 +43,17 @@ def verify_protein(query: OptimizeQuery) -> OptimizeQuery:
     # verify each residue is valid
     for aa, i in zip(query.seq, range(len(query.seq))):
         if aa.upper() not in VALID_AMINO_ACIDS:
-            raise HTTPException(400, f"Invalid residue '{aa}' in query at position: {i}")
+            raise HTTPException(
+                400, f"Invalid residue '{aa}' in query at position: {i}"
+            )
     return query
 
+
 def verify_org_id(org_id: str) -> str:
     """
     Verify that an organism id exists in the database.
     """
     org = get_species_by_id(org_id)
     if org is None:
         raise HTTPException(404, f"Organism id '{org_id}' not found in database.")
-    return org_id
+    return org_id