Merge from dev

.travis.yml

@@ -30,8 +30,8 @@ install:
   - pip install .[r]
 
 script:
-  - pytest --cov ./ test/test_*.py
+  - pytest --disable-warnings --show-capture=no --cov ./ --cov-report term --cov-report xml test/test_*.py
 
 after_success:
-  - codecov
+  - codecov -f coverage.xml
   - python-codacy-coverage -r coverage.xml

Makefile

@@ -1,15 +1,15 @@
 .DEFAULT_GOAL := pypitest
 
 test3:
-	python3 -m pytest -n 3 --disable-warnings --show-capture=no --cov ./ test/test_*.py
+	python3 -m pytest -n 3 --disable-warnings --show-capture=no --cov=ngs_toolkit --cov-report xml test/test_*.py --lf
 
 test2:
-	python2 -m pytest -n 3 --disable-warnings --show-capture=no --cov ./ test/test_*.py
+	python2 -m pytest -n 3 --disable-warnings --show-capture=no --cov=ngs_toolkit --cov-report xml test/test_*.py --lf
 
 test: test3
 
 coverage: test
-	codecov
+	codecov -f coverage.xml
 	python-codacy-coverage -r coverage.xml
 
 build: test

ngs_toolkit/atacseq.py

@@ -654,12 +654,10 @@ def get_peak_gccontent_length(
             sites = pybedtools.BedTool(bed_file)
 
         if fasta_file is None:
-            from ngs_toolkit.general import get_genome_reference
             _LOGGER.info("Reference genome FASTA file was not given, will try to get it.")
-            self._check_organism_genome(self)
             _LOGGER.info("Getting genome FASTA file for organism '{}', genome '{}'. "
                          .format(self.organism, self.genome))
-            fasta_file = get_genome_reference(organism=self.organism, genome_assembly=self.genome)
+            fasta_file = self.get_annotations(steps=['fasta'])['fasta_file']
 
         nuc = sites.nucleotide_content(fi=fasta_file).to_dataframe(comment="#")[["score", "blockStarts"]]
         nuc.columns = ["gc_content", "length"]
@@ -828,19 +826,15 @@ def get_peak_gene_annotation(self, tss_file=None, max_dist=100000):
         cols = [6, 8, 9]
 
         if tss_file is None:
-            from ngs_toolkit.general import get_tss_annotations
             _LOGGER.info("Reference TSS file was not given, will try to get TSS annotations.")
-            self._check_organism_genome(self)
             _LOGGER.info("Getting TSS annotations for organism '{}', genome '{}'. "
                          .format(self.organism, self.genome))
-            tss = pybedtools.BedTool.from_dataframe(
-                get_tss_annotations(organism=self.organism, genome_assembly=self.genome)
-                .iloc[:, list(range(6))])
-        else:
-            # extract only relevant columns
-            tss = pd.read_csv(tss_file, header=None, sep="\t")
-            tss = tss.iloc[:, list(range(6))]
-            tss = pybedtools.BedTool.from_dataframe(tss)
+            tss_file = self.get_annotations(steps=['tss'])['tss_file']
+
+        # extract only relevant columns
+        tss = pd.read_csv(tss_file, header=None, sep="\t")
+        tss = tss.iloc[:, list(range(6))]
+        tss = pybedtools.BedTool.from_dataframe(tss)
 
         if isinstance(self.sites, str):
             self.sites = pybedtools.BedTool(self.sites)
@@ -916,15 +910,14 @@ def get_peak_genomic_location(
         :returns: A dataframe with genomic context annotation for the peak set.
         """
         from ngs_toolkit.general import bed_to_index
+
         if genomic_context_file is None:
-            from ngs_toolkit.general import get_genomic_context
             _LOGGER.info("Reference genomic context file was not given, will try to get it.")
             _LOGGER.info("Getting genomic context annotations for organism '{}', genome '{}'. "
                          .format(self.organism, self.genome))
-            context = pybedtools.BedTool.from_dataframe(
-                get_genomic_context(organism=self.organism, genome_assembly=self.genome))
-        else:
-            context = pybedtools.BedTool(genomic_context_file)
+            genomic_context_file = self.get_annotations(steps=['genomic_context'])['genomic_context_file']
+
+        context = pybedtools.BedTool(genomic_context_file)
 
         if isinstance(self.sites, str):
             self.sites = pybedtools.BedTool(self.sites)
@@ -1692,9 +1685,9 @@ def characterize_regions_function(
     df[['chrom', 'start', 'end']].reset_index().to_csv(tsv_file, sep="\t", header=False, index=False)
 
     # export gene names
-    clean = df['gene_name'].str.split(",").apply(pd.Series, 1).stack().drop_duplicates()
-    clean = clean[~clean.isin(['.', 'nan'])]
-    clean.to_csv(
+    clean_gene = df['gene_name'].str.split(",").apply(pd.Series, 1).stack().drop_duplicates()
+    clean_gene = clean_gene[~clean_gene.isin(['.', 'nan', ''])]
+    clean_gene.to_csv(
             os.path.join(output_dir, "{}_genes.symbols.txt".format(prefix)),
             index=False)
     if "ensembl_gene_id" in df.columns:
@@ -1749,7 +1742,7 @@ def characterize_regions_function(
     # Enrichr
     if 'enrichr' in steps:
         _LOGGER.info("Running Enrichr for '{}'".format(prefix))
-        results = enrichr(df[['chrom', 'start', 'end', "gene_name"]])
+        results = enrichr(clean_gene.to_frame(name="gene_name"))
         results.to_csv(
             os.path.join(output_dir, "{}_regions.enrichr.csv".format(prefix)),
             index=False, encoding='utf-8')

ngs_toolkit/decorators.py

@@ -1,23 +1,34 @@
 #!/usr/bin/env python
 
 from functools import wraps
+from ngs_toolkit import _LOGGER
 
 
 def check_organism_genome(f):
-    from ngs_toolkit.general import Analysis
-
     @wraps(f)
     def wrapper(*args, **kwds):
-        Analysis._check_organism_genome(args[0])
+        attrs = ['organism', 'genome']
+        msg = "Analysis does not have 'organism' or 'genome' attributes set."
+        hint = " You can set them with analysis.set_organism_genome, for example."
+        r1 = all([hasattr(args[0], attr) for attr in attrs])
+        r2 = all([getattr(args[0], attr) is not None for attr in attrs])
+        if not all([r1, r2]):
+            _LOGGER.error(msg + hint)
+            raise AttributeError(msg)
         return f(*args, **kwds)
     return wrapper
 
 
 def check_has_sites(f):
-    from ngs_toolkit.general import Analysis
-
     @wraps(f)
     def wrapper(*args, **kwds):
-        Analysis._check_has_sites(args[0])
+        attrs = ['sites']
+        msg = "Analysis object does not have a `sites` attribute."
+        hint = " Produce one with analysis.get_consensus_sites for example."
+        r1 = all([hasattr(args[0], attr) for attr in attrs])
+        r2 = all([getattr(args[0], attr) is not None for attr in attrs])
+        if not all([r1, r2]):
+            _LOGGER.error(msg + hint)
+            raise AttributeError(msg)
         return f(*args, **kwds)
     return wrapper