Merge pull request #28 from dvklopfenstein/master

Added a new test to find error. Test can be run by cd-ing into tests and...

README.rst

@@ -21,7 +21,7 @@ Tools for Gene Ontology
          Kenta Sato (`bicycle1885 <http://github.com/bicycle1885>`_),
          Chris Mungall (`cmungall <https://github.com/cmungall>`_),
          Greg Stupp (`stuppie <https://github.com/stuppie>`_),
-         Debra Klopfenstein
+         Debra Klopfenstein(`dvklopfenstein <https://github.com/dvklopfenstein>`_)
 :Email: tanghaibao@gmail.com
 :License: BSD
 

goatools/go_enrichment.py

@@ -99,6 +99,9 @@ def run_study(self, study):
 
         pop_n, study_n = len(self.pop), len(study)
 
+        #Uncomment these to cause tests/test_enrichment_fdr.py to pass:
+        #study_count = 0
+        #pop_count   = 0
         for term, study_count in list(term_study.items()):
             pop_count = self.term_pop[term]
             p = fisher.pvalue_population(study_count, study_n,

scripts/find_enrichment.py

@@ -98,6 +98,9 @@ def check_bad_args(args):
                  "Bonferroni but slower)")
     p.add_option('--indent', dest='indent', default=False,
                  action='store_true', help="indent GO terms")
+    p.add_option('--obo', default="go-basic.obo", type="string",
+                 help="Specifies location and name of the obo file"
+                 "[default: %default]")
 
     (opts, args) = p.parse_args()
     bad = check_bad_args(args)
@@ -114,12 +117,17 @@ def check_bad_args(args):
     study_fn, pop_fn, assoc_fn = args
     study, pop = read_geneset(study_fn, pop_fn, compare=opts.compare)
     assoc = read_associations(assoc_fn)
+    #print("{}".format(assoc))
+    #print("{}".format(study))
+    #print("{}".format(pop))
 
     methods = ["bonferroni", "sidak", "holm"]
     if opts.fdr:
         methods.append("fdr")
 
-    obo_dag = GODag(obo_file="go-basic.obo")
+    obo_dag = GODag(obo_file=opts.obo)
     g = GOEnrichmentStudy(pop, assoc, obo_dag, alpha=opts.alpha,
                           study=study, methods=methods)
     g.print_summary(min_ratio=min_ratio, indent=opts.indent, pval=opts.pval)
+    print('pval={}'.format(opts.pval))
+    print('min_ratio={}'.format(min_ratio))

tests/data/study_unknown

@@ -0,0 +1,2 @@
+AT2G00000
+AT4G00001

tests/test_enrichment_fdr.py

@@ -0,0 +1,25 @@
+
+import os
+
+def test():
+  """Test to find this error below.
+
+       Traceback (most recent call last):
+         File "../scripts/find_enrichment.py", line 130, in <module>
+           study=study, methods=methods)
+         File "../scripts/../goatools/go_enrichment.py", line 93, in __init__
+           self.run_study(study)
+         File "../scripts/../goatools/go_enrichment.py", line 129, in run_study
+           p_val_distribution = calc_qval(study_count, study_n,
+       UnboundLocalError: local variable 'study_count' referenced before assignment
+  """
+  os.system("python {SCR} --alpha=0.05 {STUDY} {POP} {ASSN} --fdr --obo={OBO}".format(
+    SCR="../scripts/find_enrichment.py",
+    OBO="../go-basic.obo",
+    STUDY="data/study_unknown",
+    POP="../data/population",
+    ASSN="../data/association"))
+
+if __name__ == '__main__':
+  test()
+