UniRef90 transform

.gitignore

@@ -163,3 +163,5 @@ data/transformed/uniprot_genome_features/*.tsv
 kg_microbe/transform_utils/uniprot/tmp/relevant_file_content.txt
 kg_microbe/transform_utils/uniprot/tmp/nodes_and_edges/*
 data/transformed/uniprot_genome_features/uniprot_kgx.zip
+data/transformed/Uniref/edges.tsv
+data/transformed/Uniref/nodes.tsv

kg_microbe/transform.py

@@ -11,6 +11,7 @@
 from kg_microbe.transform_utils.rhea.rhea import RheaMappingsTransform
 from kg_microbe.transform_utils.traits.traits import TraitsTransform
 from kg_microbe.transform_utils.uniprot.uniprot import UniprotTransform
+from kg_microbe.transform_utils.uniref.uniref import UnirefTransform
 
 DATA_SOURCES = {
     # "DrugCentralTransform": DrugCentralTransform,
@@ -28,6 +29,7 @@
     "RheaMappingsTransform": RheaMappingsTransform,
     "BactoTraitsTransform": BactoTraitsTransform,
     "UniprotTransform": UniprotTransform,
+    "UnirefTransform": UnirefTransform,
 }
 
 

kg_microbe/transform_utils/constants.py

@@ -123,6 +123,8 @@
 RHEA_OLD_PREFIX = "OBO:rhea_"
 RHEA_NEW_PREFIX = "RHEA:"
 ASSAY_PREFIX = "assay:"
+UNIREF_90_PREFIX = "UniRef90:"
+
 RHEA_URI = "http://purl.obolibrary.org/obo/rhea_"
 DEBIO_OBO_PREFIX = "OBO:debio_"
 DEBIO_NEW_PREFIX = "debio:"
@@ -163,6 +165,8 @@
 NCBI_TO_SUBSTRATE_EDGE = "biolink:consumes"
 RHEA_TO_EC_EDGE = "biolink:enabled_by"
 RHEA_TO_GO_EDGE = "biolink:enables"
+NCBI_TO_CLUSTER_EDGE = "biolink:occurs_in"
+
 
 NCBI_CATEGORY = "biolink:OrganismTaxon"
 MEDIUM_CATEGORY = "biolink:ChemicalEntity"
@@ -179,6 +183,7 @@
 ATTRIBUTE_CATEGORY = "biolink:Attribute"
 METABOLITE_CATEGORY = "biolink:ChemicalEntity"
 SUBSTRATE_CATEGORY = "biolink:ChemicalEntity"
+CLUSTER_CATEGORY = "biolink:ProteinFamily"
 
 HAS_PART = "BFO:0000051"
 IS_GROWN_IN = "BAO:0002924"
@@ -194,6 +199,7 @@
 ASSESSED_ACTIVITY_RELATIONSHIP = "NCIT:C153110"
 CLOSE_MATCH = "skos:closeMatch"
 ASSOCIATED_WITH = "PATO:0001668"
+OCCURS_IN = "BFO:0000066"
 
 ID_COLUMN = "id"
 NAME_COLUMN = "name"
@@ -410,6 +416,7 @@
     GO_PREFIX,
     MEDIADIVE_MEDIUM_PREFIX,
     STRAIN_PREFIX,
+    UNIREF_90_PREFIX,
 ]
 
 HAS_PARTICIPANT_PREDICATE = "biolink:has_participant"

kg_microbe/transform_utils/uniref/__init__.py

@@ -0,0 +1,5 @@
+"""Uniref transform."""
+
+from .uniref import UnirefTransform
+
+__all__ = ["UnirefTransform"]

kg_microbe/transform_utils/uniref/uniref.py

@@ -0,0 +1,115 @@
+"""UniRef Transformation Module."""
+
+import csv
+import gc
+import os
+import sys
+from pathlib import Path
+from typing import Optional, Union
+
+from oaklib import get_adapter
+from tqdm import tqdm
+
+from kg_microbe.transform_utils.constants import (
+    CLUSTER_CATEGORY,
+    NCBI_CATEGORY,
+    NCBI_TO_CLUSTER_EDGE,
+    NCBITAXON_PREFIX,
+    OCCURS_IN,
+    UNIREF_90_PREFIX,
+)
+from kg_microbe.transform_utils.transform import Transform
+from kg_microbe.utils.dummy_tqdm import DummyTqdm
+from kg_microbe.utils.pandas_utils import drop_duplicates
+
+csv.field_size_limit(sys.maxsize - 1)  # _csv.Error: field larger than field limit (131072)
+
+
+class UnirefTransform(Transform):
+
+    """UniRef Transformation Class."""
+
+    def __init__(
+        self,
+        input_dir: Optional[Path] = None,
+        output_dir: Optional[Path] = None,
+    ):
+        """Instantiate part."""
+        source_name = "Uniref"
+        super().__init__(source_name, input_dir, output_dir)
+        self.ncbi_impl = get_adapter("sqlite:obo:ncbitaxon")
+
+    def run(self, data_file: Union[Optional[Path], Optional[str]] = None, show_status: bool = True):
+        """Run the transformation."""
+        input_file = os.path.join(
+            self.input_base_dir, "uniref90_api_subset.tsv"
+        )  # must exist already
+
+        progress_class = tqdm if show_status else DummyTqdm
+
+        with open(input_file, "r") as tsvfile, open(self.output_node_file, "w") as nodes_file, open(
+            self.output_edge_file, "w"
+        ) as edges_file:
+            # Create a CSV reader specifying the delimiter as a tab character
+            tsvreader = csv.DictReader(tsvfile, delimiter="\t")
+            node_writer = csv.writer(nodes_file, delimiter="\t")
+            edge_writer = csv.writer(edges_file, delimiter="\t")
+            source = UNIREF_90_PREFIX.strip(":")
+
+            # Write the header for the files
+            node_writer.writerow(self.node_header)
+            edge_writer.writerow(self.edge_header)
+
+            with progress_class(desc="Processing clusters...") as progress:
+                # Iterate over each row in the TSV file
+                for row in tsvreader:
+                    # Extract the desired fields
+                    cluster_id = row["Cluster ID"].replace("_", ":")
+                    cluster_name = row["Cluster Name"].lstrip("Cluster:").strip()
+                    ncbitaxon_ids = [
+                        NCBITAXON_PREFIX + x.strip() for x in row["Organism IDs"].split(";") if x
+                    ]
+                    ncbi_labels = [
+                        ncbi_label.strip()
+                        for ncbi_label in row["Organisms"].split(";")
+                        if ncbi_label
+                    ]
+                    nodes_data_to_write = [
+                        [ncbitaxon_id, NCBI_CATEGORY, ncbi_label]
+                        for ncbitaxon_id, ncbi_label in zip(
+                            ncbitaxon_ids, ncbi_labels, strict=False
+                        )
+                    ]
+                    # nodes_data_to_write.append([cluster_id, CLUSTER_CATEGORY, cluster_name])
+                    nodes_data_to_write = [
+                        sublist + [None] * (len(self.node_header) - 3)
+                        for sublist in nodes_data_to_write
+                    ]
+                    node_writer.writerows(nodes_data_to_write)
+                    gc.collect()
+
+                    # Write the cluster node
+                    cluster_node_data = [cluster_id, CLUSTER_CATEGORY, cluster_name]
+                    cluster_node_data.extend([None] * (len(self.node_header) - 3))
+                    node_writer.writerow(cluster_node_data)
+
+                    # Write the edge for the cluster
+                    edges_data_to_write = [
+                        [
+                            ncbitaxon_id,
+                            NCBI_TO_CLUSTER_EDGE,
+                            cluster_id,
+                            OCCURS_IN,
+                            source,
+                        ]
+                        for ncbitaxon_id in ncbitaxon_ids
+                    ]
+                    edge_writer.writerows(edges_data_to_write)
+                    gc.collect()
+
+                    progress.set_description(f"Processing Cluster: {cluster_id}")
+                    # After each iteration, call the update method to advance the progress bar.
+                    progress.update(2000)
+
+        drop_duplicates(self.output_node_file)
+        drop_duplicates(self.output_edge_file)

merge.yaml

@@ -105,6 +105,13 @@ merged_graph:
         filename:
           - data/transformed/uniprot_genome_features/nodes.tsv
           - data/transformed/uniprot_genome_features/edges.tsv
+    uniref:
+      input:
+        name: "Uniref"
+        format: tsv
+        filename:
+          - data/transformed/Uniref/nodes.tsv
+          - data/transformed/Uniref/edges.tsv
   operations:
     - name: kgx.graph_operations.summarize_graph.generate_graph_stats
       args: