- Update: medgen2obo.pl: (i) Abstracted adding of classes and their t…

…riples as a function, (ii) updated namespacing of classes based on what type of MedGen/UMLS identifier they are.

- Update: Namespaces MedGen, MedGen_UI (removed), MedGenCUI
- Bugfix: SSSOM metadata yaml had a typo preventing conversion
- Bugfix: Makefile: (i) needed to rename a dependency, (ii) needed to run 'analyze' step after 'stage'
- Update: Makefile: Simplified some goals

config/medgen.sssom-metadata.yml

@@ -4,8 +4,8 @@ curie_map:
   HP: http://purl.obolibrary.org/obo/HP_
   MESH: http://identifiers.org/mesh/
   MONDO: http://purl.obolibrary.org/obo/MONDO_
-  MedGen: http://purl.obolibrary.org/obo/Medgen_
-  MedGen_UID: http://purl.obolibrary.org/obo/Medgen_UID_
+  MedGen: http://purl.obolibrary.org/obo/MedGen_
+  MedGenCUI: http://purl.obolibrary.org/obo/MedGenCUI_
   NCIT: http://purl.obolibrary.org/obo/NCIT_
   OMIM: https://omim.org/entry/
   Orphanet: http://www.orpha.net/ORDO/Orphanet_

makefile

@@ -2,18 +2,17 @@
 # Running `make all` will run the full pipeline. Note that if the FTP files have already been downloaded, it'll skip
 # that part. In order to force re-download, run `make all -B`.
 .DEFAULT_GOAL := all
-.PHONY: all build stage stage-% release-artefacts analysis-artefacts clean deploy-release
+.PHONY: all build stage stage-% analyze clean deploy-release
 
 OBO=http://purl.obolibrary.org/obo
 PRODUCTS=medgen-disease-extract.obo medgen-disease-extract.owl
 TODAY ?=$(shell date +%Y-%m-%d)
 VERSION=v$(TODAY)
 
-all: build stage clean
-release-artefacts: $(PRODUCTS) medgen.sssom.tsv
-# analysis-artefacts runs more than just this file; that goal creates multiple files
-analysis-artefacts: medgen_terms_mapping_status.tsv
-build: release-artefacts analysis-artefacts
+all: build stage clean analyze
+# analyze: runs more than just this file; that goal creates multiple files
+analyze: output/medgen_terms_mapping_status.tsv
+build: $(PRODUCTS) medgen.sssom.tsv
 stage: $(patsubst %, stage-%, $(PRODUCTS))
 	mv medgen.obo output/release/
 	mv medgen.sssom.tsv output/release/
@@ -92,5 +91,6 @@ deploy-release: | output/release/
 tmp/input/mondo.sssom.tsv: | tmp/input/
 	wget http://purl.obolibrary.org/obo/mondo/mappings/mondo.sssom.tsv -O $@
 
+# creates more than just this file; that goal creates multiple files
 output/medgen_terms_mapping_status.tsv output/obsoleted_medgen_terms_in_mondo.txt: | output/
 	python src/mondo_mapping_status.py

src/medgen2obo.pl

@@ -1,6 +1,7 @@
 #!/usr/bin/perl
 use strict;
 
+# Vars
 my %th = ();
 my %rh = ();
 my %dh = ();
@@ -10,6 +11,7 @@
 
 our $PATH = "ftp.ncbi.nlm.nih.gov/pub/medgen";
 
+# Execution
 open(F,"gzip -dc $PATH/MGCONSO.RRF.gz|") || die;
 while(<F>) {
     next if m@^#@;
@@ -82,7 +84,7 @@
     chomp;
     my ($u,$c) = split(/\t/,$_);
     $uh{$c} = $u;
-    $th{$c}->{xrefs}->{"MedGen_UID:$u"} = 1;
+    $th{$c}->{xrefs}->{"MedGen:$u"} = 1;
 }
 close(F);
 
@@ -101,52 +103,12 @@
 }
 print "\n";
 
-my @ids = keys %th;
-@ids = sort @ids;
-foreach my $id (@ids) {
-    if ($id =~ /^C\d+/) {
-        # TODO: repurpose to func (this is instance 1/2)
-        my $h = $th{$id};
-        print "[Term]\n";
-        print "id: UMLS:$id\n";
-        print "name: $h->{name}\n";
-        foreach my $x (keys %{$h->{xrefs}}) {
-            $x =~ s@MSH:@MESH:@;
-            $x =~ s@NCI:@NCIT:@;
-            $x =~ s@SNOMEDCT_US:@SCTID:@;
-            print "xref: $x\n";
-        }
-        foreach (keys %{$ssh{$id} || {}}) {
-            my $ss = mk_subset($_);
-            print "subset: $ss\n";
-        }
-        foreach my $s (@{$h->{synonyms}}) {
-            my ($str, $x)= @$s;
-            $str = escq($str);
-            print "synonym: \"$str\" RELATED [$x]\n";
-        }
-        my $trelh = $rh{$id};
-        foreach my $rel (keys %{$trelh}) {
-            my $vh = $trelh->{$rel};
-            foreach my $v (keys %$vh) {
-                unless ($v eq $id) {
-                    my $tag = "relationship: $rel";
-                    if ($rel eq 'isa') {
-                        $tag = 'is_a:';
-                    }
-                    if ($rel eq 'mapped_to') {
-                        $tag = 'equivalent_to:';
-                    }
-                    print "$tag UMLS:$v {source=\"$vh->{$v}\"} ! $th{$v}->{name}\n";
-                }
-            }
-        }
-        print "\n";
-    }
-    # TODO: repurpose to func (this is instance 2/2)
+sub add_triples {
+    my ($prefix, $id) = @_;
+
     my $h = $th{$id};
     print "[Term]\n";
-    print "id: MedGen:$id\n";
+    print "id: $prefix:$id\n";
     print "name: $h->{name}\n";
     foreach my $x (keys %{$h->{xrefs}}) {
         $x =~ s@MSH:@MESH:@;
@@ -181,6 +143,18 @@
     }
     print "\n";
 }
+my @ids = keys %th;
+@ids = sort @ids;
+foreach my $id (@ids) {
+    if ($id =~ /^CN\d+/) {
+        add_triples('MedGenCUI', $id);
+    } else {
+        if ($id =~ /^C\d+/) {
+            add_triples('UMLS', $id);
+        }
+        add_triples('MedGen', $id);
+    }
+}
 
 exit 0;
 

src/mondo_mapping_status.py

@@ -12,9 +12,14 @@
 MONDO_SSSOM_TSV = INPUT_DIR / 'mondo.sssom.tsv'
 MEDGEN_SSSOM_TSV = RELEASE_OUTDIR / 'medgen.sssom.tsv'
 # MEDGEN_PREFIXES: Some of these are old, some are new, some may not be used.
-MEDGEN_PREFIXES = ['Medgen', 'MedGen', 'MEDGEN', 'Medgen_UID', 'MedGen_UID', 'UMLS', 'UMLS_CUI']
+# todo: If I couldn't convert SSSOM properly with MedGen_CUI, souldn't UMLS_CUI have a problem? though i think it's just coming from previous work in mondo maybe. it's not being used in this ingest
+MEDGEN_PREFIXES = [
+    'Medgen', 'MedGen', 'MEDGEN', 'MedGenCUI', 'UMLS', 'UMLS_CUI',
+    # 'Medgen_UID', 'MedGen_UID', 'Medgen_CUI', 'MedGen_CUI', 'Medgen_CUI'
+]
 CURIE = str
 
+# TODO: Mappings can be considered skos:exactMatch
 
 def ids_prefixless(ids: Set[str]) -> Set[str]:
     """Remove prefix"""
@@ -92,16 +97,19 @@ def medgen_mondo_mapping_status(mondo_predicate_filter: List[str] = None):
     file_suffix = '' if not mondo_predicate_filter \
         else '-mondo-exacts-only' if mondo_predicate_filter == ['skos:exactMatch'] \
         else '-custom'
+
     # Read sources
     medgen_all_ids, medgen_in_medgen, medgen_in_mondo = \
         read_mapping_sources(mondo_predicate_filter=mondo_predicate_filter)
+
     # Special operations
     # - Inconsistent prefixes between what Mondo used before and will going forward. In this case, stripping prefixes
     # should be OK, at least for now.
     medgen_all_ids = ids_prefixless(medgen_all_ids)
     medgen_in_medgen = ids_prefixless(medgen_in_medgen)
     medgen_in_mondo = ids_prefixless(medgen_in_mondo)
-    # Report
+
+    # Generate reports
     report_obs_medgen_in_mondo(medgen_in_mondo, medgen_in_medgen)
     report_existing_overlap(medgen_all_ids, medgen_in_medgen, medgen_in_mondo, file_suffix)