Merge pull request #12 from monarch-initiative/ordo-mappings2

ICD11 --> ORDO mappings

.gitignore

@@ -130,3 +130,6 @@ dmypy.json
 
 .idea/
 tmp/
+
+# Custom
+_archive/

README.md

@@ -1,11 +1,14 @@
-# ICD11 Foundation Ingest
-
-## Pre-reqs
-- Docker
-- Docker images  
-  One or both of the following, depending on if you want to run the stable build `latest` or `dev`:
-  - a. `docker pull obolibrary/odkfull:latest`
-  - b. `docker pull obolibrary/odkfull:dev`
-
-## Running
-`sh run.sh make all`
+# ICD11 Foundation Ingest
+
+## Pre-reqs
+- Docker
+- Docker images  
+  One or both of the following, depending on if you want to run the stable build `latest` or `dev`:
+  - a. `docker pull obolibrary/odkfull:latest`
+  - b. `docker pull obolibrary/odkfull:dev`
+
+## Local development setup
+`pip install -r requirements.txt`
+
+## Running
+`sh run.sh make all`

config/icd11.sssom-metadata.yml

@@ -0,0 +1,15 @@
+creator_id: orcid:0000-0002-2906-7319
+curie_map:
+  icd11.foundation: http://id.who.int/icd/entity/
+  MONDO: http://purl.obolibrary.org/obo/MONDO_
+  oboInOwl: http://www.geneontology.org/formats/oboInOwl#
+  orcid: https://orcid.org/
+  owl: http://www.w3.org/2002/07/owl#
+  Orphanet: http://www.orpha.net/ORDO/Orphanet_
+  rdf: http://www.w3.org/1999/02/22-rdf-syntax-ns#
+  rdfs: http://www.w3.org/2000/01/rdf-schema#
+  semapv: https://w3id.org/semapv/
+  skos: http://www.w3.org/2004/02/skos/core#
+  sssom: https://w3id.org/sssom/
+license: http://w3id.org/sssom/license/unspecified
+mapping_provider: https://www.orpha.net/

makefile

@@ -1,12 +1,20 @@
 .DEFAULT_GOAL := all
-.PHONY: all release clean
+.PHONY: all release clean ontology mappings mappings-validate help
 TODAY ?=$(shell date +%Y-%m-%d)
 VERSION=v$(TODAY)
 SOURCE_URL=https://icd11files.blob.core.windows.net/tmp/whofic-2023-04-08.owl.gz
 
 
 # MAIN COMMANDS / GOALS ------------------------------------------------------------------------------------------------
-all: tmp/output/release/icd11foundation.owl
+all: ontology mappings
+
+ontology: tmp/output/release/icd11foundation.owl
+
+mappings: mappings-validate
+
+mappings-validate: tmp/output/release/mondo_exactmatch_icd11foundation.sssom.tsv
+	sssom validate tmp/output/release/mondo_exactmatch_icd11foundation.sssom.tsv
+	sssom validate tmp/output/release/ordo_exactmatch_icd11foundation.sssom.tsv
 
 clean:
 	rm -rf tmp/
@@ -26,6 +34,7 @@ tmp/input/:
 
 tmp/input/source.owl: tmp/input/source.gz
 	gunzip -c $< > $@
+	rm $<
 
 tmp/input/source.gz: | tmp/input/
 	wget ${SOURCE_URL} -O $@
@@ -35,6 +44,37 @@ release: | tmp/output/release/
 	@test $(VERSION)
 	gh release create $(VERSION) --notes "New release." --title "$(VERSION)" tmp/output/release/*
 
+# Mappings
+# todo: I'd like to do USE_PREBUILT=false, but this error is occuring:
+#  cp: cannot stat 'tmp/input/mondo/src/ontology/mappings/mondo.sssom.tsv': No such file or directory
+#  ...even though I checked and the file at that path does exist.
+USE_PREBUILT=true
+
+tmp/input/mondo/:
+	rm -rf $@ &&\
+	cd tmp/input/ &&\
+	git clone --depth 1 https://github.com/monarch-initiative/mondo
+
+tmp/input/mondo.sssom.tsv: tmp/input/mondo/
+	if [ $(USE_PREBUILT) = true ]; then wget https://raw.githubusercontent.com/monarch-initiative/mondo/master/src/ontology/mappings/mondo.sssom.tsv -O $@; else cd tmp/input/ &&\
+		cd mondo/src/ontology &&\
+		make mondo.owl mappings -B MIR=false IMP=false MIR=false &&\
+		cp tmp/input/mondo/src/ontology/mappings/mondo.sssom.tsv $@; fi
+
+# todo: Stable URI/filename issue: https://github.com/monarch-initiative/icd11/pull/12#discussion_r1542187711
+tmp/input/Orphanet_Nomenclature_Pack_EN/ORPHA_ICD11_mapping_en_newversion_2023.xml: | tmp/input/
+	wget https://www.orphadata.com/data/nomenclature/packs/Orphanet_Nomenclature_Pack_EN.zip -O tmp/input/Orphanet_Nomenclature_Pack_EN.zip
+	unzip tmp/input/Orphanet_Nomenclature_Pack_EN.zip -d tmp/input/Orphanet_Nomenclature_Pack_EN
+
+# todo: rename icd11.sssom-metadata.yml --> icd11foundation.sssom-metadata.yml
+tmp/output/release/mondo_exactmatch_icd11foundation.sssom.tsv tmp/output/release/ordo_exactmatch_icd11foundation.sssom.tsv: tmp/input/Orphanet_Nomenclature_Pack_EN/ORPHA_ICD11_mapping_en_newversion_2023.xml tmp/input/mondo.sssom.tsv | tmp/output/release/
+	python3 src/mappings.py \
+	--input-nomenclature-xml tmp/input/Orphanet_Nomenclature_Pack_EN/ORPHA_ICD11_mapping_en_newversion_2023.xml \
+	--input-sssom-config config/icd11.sssom-metadata.yml \
+	--input-mondo-sssom tmp/input/mondo.sssom.tsv \
+	--outpath-ordo-mappings tmp/output/release/ordo_exactmatch_icd11foundation.sssom.tsv \
+	--outpath-mondo-mappings tmp/output/release/mondo_exactmatch_icd11foundation.sssom.tsv
+
 # HELP -----------------------------------------------------------------------------------------------------------------
 help:
 	@echo "-----------------------------------"

requirements-unlocked.txt

@@ -0,0 +1,3 @@
+pandas
+pyyaml
+sssom

requirements.txt

@@ -0,0 +1,60 @@
+annotated-types==0.6.0
+attrs==23.2.0
+certifi==2024.2.2
+charset-normalizer==3.3.2
+click==8.1.7
+curies==0.7.9
+Deprecated==1.2.14
+deprecation==2.1.0
+distlib==0.3.8
+exceptiongroup==1.2.0
+filelock==3.13.1
+hbreader==0.9.1
+idna==3.6
+importlib_resources==6.4.0
+iniconfig==2.0.0
+isodate==0.6.1
+json-flattener==0.1.9
+jsonasobj2==1.0.4
+jsonschema==4.21.1
+jsonschema-specifications==2023.12.1
+linkml-runtime==1.7.5
+networkx==3.3
+numpy==1.26.4
+packaging==24.0
+pandas==2.2.1
+pansql==0.0.1
+pbr==6.0.0
+platformdirs==4.2.0
+pluggy==1.4.0
+prefixcommons==0.1.12
+prefixmaps==0.2.3
+pydantic==2.6.4
+pydantic_core==2.16.3
+pyparsing==3.1.2
+pytest==8.1.1
+pytest-logging==2015.11.4
+python-dateutil==2.9.0.post0
+PyTrie==0.4.0
+pytz==2024.1
+PyYAML==6.0.1
+rdflib==7.0.0
+referencing==0.34.0
+requests==2.31.0
+rpds-py==0.18.0
+scipy==1.13.0
+six==1.16.0
+sortedcontainers==2.4.0
+SPARQLWrapper==2.0.0
+SQLAlchemy==2.0.29
+sssom==0.4.6
+sssom-schema==0.15.2
+stevedore==5.1.0
+tomli==2.0.1
+typing_extensions==4.11.0
+tzdata==2024.1
+urllib3==2.2.1
+validators==0.28.0
+virtualenv==20.25.0
+virtualenv-clone==0.5.7
+wrapt==1.16.0

run.sh

@@ -14,8 +14,6 @@
 #
 # See README-editors.md for more details.
 
-set -e
-
 if [ -f run.sh.conf ]; then
     . ./run.sh.conf
 fi
@@ -31,22 +29,6 @@ elif [ -f "$HOME/Library/Application Support/ontology-development-kit/github/tok
     GH_TOKEN=$(cat "$HOME/Library/Application Support/ontology-development-kit/github/token")
 fi
 
-# SSH agent socket
-# On macOS, we cannot use $SSH_AUTH_SOCK directly,
-# we need to use a "magic" socket instead.
-case "$(uname)" in
-Darwin)
-    ODK_SSH_AUTH_SOCKET=/run/host-services/ssh-auth.sock
-    ;;
-*)
-    ODK_SSH_AUTH_SOCKET=$SSH_AUTH_SOCK
-    ;;
-esac
-ODK_SSH_BIND=
-if [ -n "$ODK_SSH_AUTH_SOCKET" ]; then
-    ODK_SSH_BIND=",$ODK_SSH_AUTH_SOCKET:/run/host-services/ssh-auth.sock"
-fi
-
 ODK_IMAGE=${ODK_IMAGE:-odkfull}
 TAG_IN_IMAGE=$(echo $ODK_IMAGE | awk -F':' '{ print $2 }')
 if [ -n "$TAG_IN_IMAGE" ]; then
@@ -58,9 +40,6 @@ ODK_TAG=${ODK_TAG:-latest}
 ODK_JAVA_OPTS=${ODK_JAVA_OPTS:--Xmx20G}
 ODK_DEBUG=${ODK_DEBUG:-no}
 
-ODK_USER_ID=${ODK_USER_ID:-$(id -u)}
-ODK_GROUP_ID=${ODK_GROUP_ID:-$(id -g)}
-
 # Convert OWLAPI_* environment variables to the OWLAPI as Java options
 # See http://owlcs.github.io/owlapi/apidocs_4/org/semanticweb/owlapi/model/parameters/ConfigurationOptions.html
 # for a list of allowed options
@@ -77,26 +56,25 @@ if [ x$ODK_DEBUG = xyes ]; then
     echo "Running ${IMAGE} with ${ODK_JAVA_OPTS} of memory for ROBOT and Java-based pipeline steps."
     TIMECMD="/usr/bin/time -f ### DEBUG STATS ###\nElapsed time: %E\nPeak memory: %M kb"
 fi
-rm -f tmp/debug.log
 
-VOLUME_BIND=$PWD/../../:/work$ODK_SSH_BIND
-WORK_DIR=/work/src/ontology
+VOLUME_BIND=$PWD:/work
+WORK_DIR=/work
 
 if [ -n "$ODK_BINDS" ]; then
     VOLUME_BIND="$VOLUME_BIND,$ODK_BINDS"
 fi
 
 if [ -n "$USE_SINGULARITY" ]; then
-    
+
     singularity exec --cleanenv $ODK_SINGULARITY_OPTIONS \
-        --env "ROBOT_JAVA_ARGS=$ODK_JAVA_OPTS,JAVA_OPTS=$ODK_JAVA_OPTS,SSH_AUTH_SOCK=/run/host-services/ssh-auth.sock,ODK_USER_ID=$ODK_USER_ID,ODK_GROUP_ID=$ODK_GROUP_ID,ODK_DEBUG=$ODK_DEBUG" \
+        --env "ROBOT_JAVA_ARGS=$ODK_JAVA_OPTS,JAVA_OPTS=$ODK_JAVA_OPTS" \
         --bind $VOLUME_BIND \
         -W $WORK_DIR \
         docker://obolibrary/$ODK_IMAGE:$ODK_TAG $TIMECMD "$@"
 else
     BIND_OPTIONS="-v $(echo $VOLUME_BIND | sed 's/,/ -v /')"
     docker run $ODK_DOCKER_OPTIONS $BIND_OPTIONS -w $WORK_DIR \
-        -e ROBOT_JAVA_ARGS="$ODK_JAVA_OPTS" -e JAVA_OPTS="$ODK_JAVA_OPTS" -e SSH_AUTH_SOCK=/run/host-services/ssh-auth.sock -e ODK_USER_ID=$ODK_USER_ID -e ODK_GROUP_ID=$ODK_GROUP_ID -e ODK_DEBUG=$ODK_DEBUG \
+        -e ROBOT_JAVA_ARGS="$ODK_JAVA_OPTS" -e JAVA_OPTS="$ODK_JAVA_OPTS" \
         --rm -ti obolibrary/$ODK_IMAGE:$ODK_TAG $TIMECMD "$@"
 fi