Panel design notebooks

CHANGELOG.md

@@ -4,6 +4,9 @@ This project adheres to [Semantic Versioning](http://semver.org/).
 
 ## [Unreleased]
 
+### Added
+- Panel design notebooks (#157).
+
 ### Fixed
 - Debugged a test that counts observed haplotypes (#154).
 - Replaced global pooled Ae values with 26-population average as the default Ae reported (#155, #158).

dbbuild/repeats.py

@@ -58,6 +58,12 @@ def parse_ucsc_rmsk_track(path):
         "id",
     ]
     table = pd.read_csv(path, sep="\t", names=header)
+    table = table[
+        (~table.repClass.isin(("SINE", "LINE", "LTR")))
+        | ((table.repClass == "SINE") & (table.swScore > 929))
+        | ((table.repClass == "LINE") & (table.swScore > 411))
+        | ((table.repClass == "LTR")  & (table.swScore > 909))
+    ]
     return table.groupby("genoName")
 
 

microhapdb/__init__.py

@@ -11,7 +11,7 @@
 # -------------------------------------------------------------------------------------------------
 
 from . import nomenclature
-from .tables import markers, merged, populations, frequencies, repeats, indels, variantmap, hg38
+from .tables import markers, merged, populations, frequencies, indels, variantmap, hg38
 from .population import Population
 from .marker import Marker, Locus
 from microhapdb import cli

microhapdb/cli/marker.py

@@ -17,7 +17,6 @@
 import pandas as pd
 import sys
 from textwrap import dedent
-from warnings import warn
 
 
 def main(args):
@@ -96,7 +95,7 @@ def display(
             for marker in markers:
                 loci[marker.locus].markers.append(marker)
             table = pd.concat([locus.definition for locus in loci.values()])
-            table = table.rename(columns={"ChromOffset": f"OffsetHg38"})
+            table = table.rename(columns={"ChromOffset": "OffsetHg38"})
             table.to_csv(sys.stdout, sep="\t", index=False)
         else:
             raise ValueError(f'unsupported view format "{view_format}"')