Merge pull request #29 from sbslee/0.8.0-dev

0.8.0 dev

Author: sbslee

CHANGELOG.rst

@@ -1,6 +1,14 @@
 Changelog
 *********
 
+0.8.0 (2021-11-20)
+------------------
+
+* Update :meth:`api.core.sort_alleles` method to also sort alleles by name for genes that do not use the star allele nomenclature (e.g. the DPYD gene).
+* Add new method :meth:`api.core.is_legit_allele`.
+* Update :meth:`api.core.predict_phenotype` method to first check if the two alleles are legit.
+* Add new genes: ABCB1, CYP1A1, CYP1B1, CYP4A11, CYP4A22, CYP4B1, CYP17A1, CYP19A1, G6PD, IFNL3, POR, PTGIS, SLCO1B3, SULT1A1, TBXAS1, UGT1A4, XPC.
+
 0.7.0 (2021-10-23)
 ------------------
 

README.rst

@@ -185,7 +185,7 @@ For getting help on the CLI:
        estimate-phase-beagle
                            Estimate haplotype phase of observed variants with the Beagle program.
        filter-samples      Filter Archive file for specified samples.
-       import-read-depth   Import read depth data for target gene.
+       import-read-depth   Import read depth data for the target gene.
        import-variants     Import variant data for the target gene.
        plot-bam-copy-number
                            Plot copy number profile from CovFrame[CopyNumber].
@@ -232,6 +232,13 @@ For getting help on a specific submodule (e.g. utils):
    >>> from pypgx.api import utils
    >>> help(utils)
 
+For getting help on a specific method (e.g. predict_phenotype):
+
+.. code:: python3
+
+   >>> import pypgx
+   >>> help(pypgx.predict_phenotype)
+
 CLI examples
 ============
 

docs/cli.rst

@@ -36,7 +36,7 @@ For getting help on the CLI:
        estimate-phase-beagle
                            Estimate haplotype phase of observed variants with the Beagle program.
        filter-samples      Filter Archive file for specified samples.
-       import-read-depth   Import read depth data for target gene.
+       import-read-depth   Import read depth data for the target gene.
        import-variants     Import variant data for the target gene.
        plot-bam-copy-number
                            Plot copy number profile from CovFrame[CopyNumber].
@@ -768,7 +768,7 @@ train-cnv-caller
    Train a CNV caller for the target gene.
    
    This command will return a SVM-based multiclass classifier that makes CNV
-   calls using the one-vs-rest stategy.
+   calls using the one-vs-rest strategy.
    
    Positional arguments:
      copy-number           Archive file with the semantic type 

docs/create.py

@@ -211,6 +211,13 @@
    >>> from pypgx.api import utils
    >>> help(utils)
 
+For getting help on a specific method (e.g. predict_phenotype):
+
+.. code:: python3
+
+   >>> import pypgx
+   >>> help(pypgx.predict_phenotype)
+
 CLI examples
 ============
 

pypgx/__init__.py

@@ -3,6 +3,7 @@
     collapse_alleles,
     has_phenotype,
     has_score,
+    is_legit_allele,
     is_target_gene,
     get_default_allele,
     get_function,

pypgx/api/core.py

@@ -14,13 +14,18 @@
 
 FUNCTION_ORDER = [
     'No Function',
+    'Severely Decreased Function',
     'Decreased Function',
     'Possible Decreased Function',
     'Increased Function',
     'Possible Increased Function',
+    'Class I (Deficient with CNSHA)',
+    'Class II (Deficient)',
+    'Class III (Deficient)',
     'Uncertain Function',
     'Unknown Function',
     'Normal Function',
+    'Class IV (Normal)',
 ]
 
 class AlleleNotFoundError(Exception):
@@ -202,6 +207,24 @@ def has_score(gene):
 
     return gene in df[df.PhenotypeMethod == 'Score'].Gene.unique()
 
+def is_legit_allele(gene, allele):
+    """
+    Return True if specified allele exists in the allele table.
+
+    Parameters
+    ----------
+    gene : str
+        Target gene.
+    allele : str
+        Allele to be tested.
+
+    Returns
+    -------
+    bool
+        True if the allele is legit.
+    """
+    return allele in list_alleles(gene)
+
 def is_target_gene(gene):
     """
     Return True if specified gene is one of the target genes.
@@ -465,7 +488,7 @@ def get_score(gene, allele):
     df = df[(df.Gene == gene) & (df.StarAllele == allele)]
 
     if df.empty:
-        raise AlleleNotFoundError(gene + allele)
+        raise AlleleNotFoundError(gene, allele)
 
     return df.ActivityScore.values[0]
 
@@ -960,6 +983,10 @@ def one_row(r, score):
     elif phenotype_method == 'Diplotype':
         df = load_diplotype_table()
         df = df[df.Gene == gene]
+        if not is_legit_allele(gene, a):
+            raise AlleleNotFoundError(gene, a)
+        if not is_legit_allele(gene, b):
+            raise AlleleNotFoundError(gene, b)
         l = [f'{a}/{b}', f'{b}/{a}']
         i = df.Diplotype.isin(l)
         try:
@@ -1095,7 +1122,7 @@ def sort_alleles(
 
     >>> alleles = ['*1', '*2', '*4', '*10']
 
-    We can sort them by their prioirty with ``method='priority'``:
+    We can sort the alleles by their prioirty with ``method='priority'``:
 
     >>> import pypgx
     >>> alleles = pypgx.sort_alleles(alleles, by='priority', gene='CYP2D6', assembly='GRCh37')
@@ -1107,8 +1134,17 @@ def sort_alleles(
     >>> alleles = pypgx.sort_alleles(alleles, by='name')
     >>> alleles
     ['*1', '*2', '*4', '*10']
+
+    Note that we can also sort alleles by name for genes that do not use the
+    star allele nomenclature (e.g. the *DPYD* gene):
+
+    >>> alleles = ['c.557A>G', 'c.2194G>A (*6)', 'c.496A>G', 'Reference', 'c.1627A>G (*5)']
+    >>> pypgx.sort_alleles(alleles, by='name')
+    ['Reference', 'c.496A>G', 'c.557A>G', 'c.1627A>G (*5)', 'c.2194G>A (*6)']
     """
     def func1(allele):
+        if gene is None:
+            raise ValueError('Gene is required when sorting by priority')
         if not is_target_gene(gene):
             raise NotTargetGeneError(gene)
         function = get_function(gene, allele)
@@ -1122,13 +1158,18 @@ def func1(allele):
         return (a, b, c, d)
 
     def func2(allele):
+        n = 99999
         cn = 1
-        if '*' not in allele:
-            n = 999
+        if allele == 'Reference':
+            n = 0
+        elif 'c.' in allele: # For the DPYD gene
+            n = int(''.join([x for x in allele.split('>')[0] if x.isdigit()]))
+        elif '*' not in allele:
+            pass
         else:
             _ = allele.split('+')[0].split('x')[0].replace('*', '')
             if not _[0].isdigit():
-                n = 999
+                pass
             else:
                 n = int(''.join([x for x in _ if x.isdigit()]))
             if 'x' in allele.split('+')[0]: