AOP-tk wrappers for download, chemical identification + matching + normalization, PDF parsing, relationship finding

tools/aoptk/.shed.yml

@@ -4,13 +4,13 @@ remote_repository_url: "https://github.com/rdurnik/aoptk"
 homepage_url: "https://github.com/rdurnik/aoptk"
 categories:
   - Machine Learning
-description: "AOP-toolkit (aoptk) is a Python package designed to support the development of Adverse Outcome Pathways (AOPs) that require extensive data mining."
+description: "AOP-toolkit (aoptk) is a Python package designed to support data mining and analysis of toxicological outcomes."
 long_description: |
-  "AOP-toolkit (aoptk) is a Python package developed to support the construction of Adverse Outcome Pathways (AOPs) that require extensive mining and integration of toxicological data from heterogeneous sources. It enables researchers to collect literature from databases such as PubMed and Europe PMC, extract relevant information from full-text publications, and analyze complex, unstructured data using large language models. The toolkit also provides functionality for normalizing chemical names across publications, helping ensure consistency and interoperability."
+  "AOP-toolkit (aoptk) is a Python package for mining and analyzing toxicological and biomedical literature. Originally developed to support the construction of Adverse Outcome Pathways (AOPs), it provides general-purpose tools for retrieving, processing, and analyzing scientific publications."
 auto_tool_repositories:
   name_template: "{{ tool_id }}"
   description_template: "{{ tool_name }} tool from the aoptk package"
 suite:
   name: suite_aoptk
-  description: AOP-toolkit (aoptk) is a Python package developed to support the construction of Adverse Outcome Pathways (AOPs) that require extensive mining and integration of toxicological data from heterogeneous sources.
+  description: AOP-toolkit (aoptk) is a Python package for mining and analyzing toxicological and biomedical literature. Originally developed to support the construction of Adverse Outcome Pathways (AOPs), it provides general-purpose tools for retrieving, processing, and analyzing scientific publications.
   type: repository_suite_definition

tools/aoptk/aoptk_chemical_identifier.xml

@@ -1,11 +1,12 @@
-<tool id="aoptk_chemical_identifier" name="aoptk chemical identifier" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="25.1" license="MIT">
+<tool id="aoptk_chemical_identifier" name="aoptk chemical identifier" version="0.2.0+galaxy0" profile="25.1" license="MIT">
     <description>Detect chemicals in scientific literature and filter based on a screening list.</description>
     <macros>
         <import>macros.xml</import>
     </macros>
 
     <requirements>
-        <expand macro="requirements"/>
+        <requirement type="package" version="0.2.0">aoptk</requirement>
+        <expand macro="email_credentials"/>
     </requirements>
 
     <command detect_errors="exit_code"><![CDATA[
@@ -18,7 +19,11 @@
     ]]></command>
 
     <inputs>
-        <expand macro="inputs"/>
+        <expand macro="query"/>
+        <param name="literature_database" type="select" label="Select which database to search." help="PubMed only accepts searches with up to 10 000 results. No limit for Europe PMC and PMC.">
+            <option value="europepmc" selected="true">Europe PMC</option>
+            <option value="pubmed">PubMed</option>
+        </param>
         <param argument="--chemical_database" type="data" format="xlsx" label="Chemical database" help="Custom chemical database with toxicologically relevant chemicals. Excel file with single column: chemical_name. Examples can be found in Citations." />
     </inputs>
 
@@ -96,5 +101,6 @@ For the purposes of this publication, Chemical-Identifier was originally develop
     <citations>
         <citation type="doi">10.18653/v1/w19-5034</citation>
         <citation type="doi">10.1093/bioinformatics/btp163</citation>
+        <citation type="doi">10.5281/zenodo.20036704</citation>
     </citations>
 </tool>

tools/aoptk/aoptk_download_publication_data.xml

@@ -0,0 +1,148 @@
+<tool id="aoptk_download_publication_data" name="aoptk download publication data" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="25.1" license="MIT">
+    <description>Download publication data - full text publications and figures, abstracts, or PDFs.</description>
+    <macros>
+        <import>macros.xml</import>
+    </macros>
+
+    <requirements>
+        <expand macro="requirements"/>
+    </requirements>
+
+    <command detect_errors="exit_code"><![CDATA[
+        python3 '${download_publication_data}'
+    ]]></command>
+
+<configfiles>
+<configfile name="download_publication_data">
+import os
+from Bio import Entrez
+
+with open("$input_file", "r") as f:
+    ids = [line.strip() for line in f.readlines()]
+    email = os.environ.get("EMAIL")
+    api_key = os.environ.get("NCBI_API_KEY")
+
+
+#if $literature.database == "pmc":
+from aoptk.literature.databases.pmc import PMC
+Entrez.email = email
+Entrez.api_key = api_key
+database = PMC.__new__(PMC)
+database.storage = "./"
+database.figure_storage = "./figures"
+database.id_list = ids
+
+#elif $literature.database == "europepmc":
+from aoptk.literature.databases.europepmc import EuropePMC
+database = EuropePMC("", storage = "./", figure_storage="./figures")
+database.id_list = ids
+
+#elif $literature.database == "pubmed":
+from aoptk.literature.databases.pubmed import PubMed
+Entrez.email = email
+Entrez.api_key = api_key 
+database = PubMed.__new__(PubMed)
+database.id_list = ids
+
+#else:
+raise ValueError("Select valid database.")
+#end if
+
+
+#if $literature.type_choice == "publications":
+publications = database.get_publications()
+for publication in publications:
+    print(publication.figures)
+    with open(f"{publication.id}.txt", "w") as f:
+        f.write(publication.full_text)
+
+#elif $literature.type_choice == "abstracts":
+abstracts = database.get_abstracts()
+for abstract in abstracts:
+    with open(f"{abstract.id}.txt", "w") as f:
+        f.write(abstract.text)
+
+#elif $literature.type_choice == "pdfs":
+pdfs = database.get_pdfs()
+
+#else:
+raise ValueError("Select valid data type.")
+#end if
+
+
+</configfile>
+</configfiles>
+
+    <inputs>
+        <param name="input_file" type="data" format="txt" label="List of IDs to search for." help="Input text file with IDs to search for." />
+        <conditional name="literature">
+            <param name="database" type="select" label="Select which database to search." help="PubMed only accepts searches with up to 10 000 results. No limit for Europe PMC and PMC.">
+                <option value="pmc" selected="true">PMC</option>
+                <option value="pubmed">PubMed</option>
+                <option value="europepmc">Europe PMC</option>
+            </param>
+            <when value="pubmed">
+                <param name="type_choice" type="select" label="Data type">
+                    <option value="abstracts">Abstracts</option>
+                </param>
+            </when>
+            <when value="europepmc">
+                <param name="type_choice" type="select" label="Data type">
+                    <option value="publications">Full text publications</option>
+                    <option value="abstracts">Abstracts</option>
+                </param>
+            </when>
+            <when value="pmc">
+                <param name="type_choice" type="select" label="Data type">
+                    <option value="publications">Full text publications</option>
+                    <option value="pdfs">PDFs</option>
+                </param>
+            </when>
+        </conditional>
+    </inputs>
+
+    <outputs>
+        <collection name="text" type="list" label="Downloaded ${database} ${type_choice}" format="txt">
+            <filter>literature['type_choice'] == 'publications' or literature['type_choice'] == 'abstracts'</filter>
+            <discover_datasets pattern="(?P&lt;designation&gt;.*)\.txt$" ext="txt" />
+        </collection>
+        <collection name="pdfs" type="list" label="Downloaded ${database} PDFs" format="pdf">
+            <filter>literature['type_choice'] == 'pdfs'</filter>
+            <discover_datasets pattern="(?P&lt;designation&gt;.*)\.pdf$" ext="pdf" />
+        </collection>
+    </outputs>
+
+    <tests>
+        <test expect_num_outputs="1">
+            <param name="literature|database" value="pmc"/>
+            <param name="literature|type_choice" value="publications"/>
+            <param name="input_file" value="test-data/ids.txt"/>
+            <output_collection name="text" type="list" count="2"/>
+        </test>
+        <test expect_num_outputs="1">
+            <param name="literature|database" value="europepmc"/>
+            <param name="literature|type_choice" value="abstracts"/>
+            <param name="input_file" value="test-data/ids.txt"/>
+            <output_collection name="text" type="list" count="2"/>
+        </test>
+        <test expect_num_outputs="1">
+            <param name="literature|database" value="pmc"/>
+            <param name="literature|type_choice" value="pdfs"/>
+            <param name="input_file" value="test-data/ids.txt"/>
+            <output_collection name="pdfs" type="list" count="2"/>
+        </test>
+    </tests>
+
+    <help><![CDATA[
+Download Publication Data
+=========================
+
+Tools to download publication data - full text publications (including abstracts) and figures, abstracts, or PDFs.
+
+    ]]></help>
+
+    <citations>
+        <citation type="doi">10.5281/zenodo.20036704</citation>
+    </citations>
+
+</tool>

tools/aoptk/aoptk_finalize_output.xml

@@ -0,0 +1,74 @@
+<tool id="aoptk_final_modification" name="aoptk final modification" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="25.1" license="MIT">
+    <description>Finalize the output of relationship identification.</description>
+    <macros>
+        <import>macros.xml</import>
+    </macros>
+
+    <requirements>
+        <expand macro="requirements"/>
+        <requirement type="package" version="3.1.5">openpyxl</requirement>
+    </requirements>
+
+    <command detect_errors="exit_code"><![CDATA[
+        python3 '${final_modification}'
+    ]]></command>
+
+<configfiles>
+<configfile name="final_modification">
+import pandas as pd
+
+def join_unique(series):
+    values = series.dropna().astype(str).str.strip()
+    return ";".join(pd.unique(values))
+
+output = pd.read_csv("$input_file", sep="\t")
+
+file_name = output["effect"].unique()[0].strip().lower().replace(" ", "_")
+
+output["relationship"] = output["relationship"].astype(str).str.strip().str.lower()
+relationships = sorted(output["relationship"].unique())
+final_output = output[["heading"]].drop_duplicates()
+
+for relationship in relationships:
+    subset = output[output["relationship"] == relationship]
+    result = subset.groupby("heading", as_index=False, dropna=False).agg(
+        hits=("id", "size"),
+        ids=("id", join_unique),
+    )
+    result.rename(columns={"hits": f"{relationship}", "ids": f"{relationship} IDs"}, inplace=True)
+    final_output = final_output.merge(result, on="heading", how="left")
+
+final_output.to_excel("./final_output.xlsx", index=False)
+
+
+</configfile>
+</configfiles>
+
+    <inputs>
+        <param name="input_file" type="data" format="tabular" label="TSV with heading column." help="Input tsv file with chemical column." />
+    </inputs>
+
+    <outputs>
+        <data name="final_output" format="xlsx" from_work_dir="final_output.xlsx" label="Final output." />
+    </outputs>
+
+    <tests>
+        <test>
+           <param name="input_file" value="test-data/chemicals.tsv"/>
+           <output name="final_output" file="test-data/final_output.xlsx" compare="sim_size" delta="100"/>
+        </test>
+    </tests>
+
+    <help><![CDATA[
+Final Modification
+===================
+
+Tool to remove duplicates and finalize the output of relationship identification.
+
+    ]]></help>
+
+    <citations>
+        <citation type="doi">10.5281/zenodo.20036704</citation>
+    </citations>
+
+</tool>