Project structure and data

This page describes the project folder structure, and the input and output data.

Project folder organisation

We recommend creating an empty folder for the project. The path of this project will be the input of the API. The input folder needs to be manually created and populated with the data; the rest of the folders will be generated by the API.

PROJECT_FOLDER
├── input
│   └── ...
└── output
    ├── domain_ontology
    │   ├── ...
    ├── intermediate
    │   ├── ...
    └── report
        └── ...

---

Data schema

General points:

• default_id column is mandatory (nodes and nodes_obsolete tables).

• default_label column is optional in all tables.

• source_id and target_id must be the same format as default_id,

  for more on this please read node ID schema,

Node ID schema

All nodes must be formatted as follows NAMESPACE:CODE, for example: MONDO:0000123, SNOMED:1233425, where

• NAMESPACE: denotes the source of the node, e.g. MONDO, SNOMED;

  this should be all caps for uniformity.

• CODE: is the actual node identifier code that must not contain :.

• NAMESPACE and CODE parts must be concatenated by the : character.

• There must be only one : character in the node ID.

---

Input data

The input requires four tables with the below specified schema. All of the tables are required to be present, however nodes_obsolete.csv can be an empty table with a header.

PROJECT_FOLDER
├── input
│   ├── config.json
│   ├── edges_hierarchy.csv
│   ├── mappings.csv
│   ├── nodes.csv
│   └── nodes_obsolete.csv
├── ...

Table nodes.csv

default_id	default_label
MONDO:0004979	asthma

Table nodes_obsolete.csv

default_id	default_label
MONDO:0006775	obsolete haemophilus influenzae meningitis

Table edges_hierarchy.csv

source_id	target_id	relation	prov
MONDO:0008798	MONDO:0019211	rdfs:subClassOf	MONDO

Table mappings.csv

source_id	target_id	relation	prov
MONDO:0004979	SNOMED:31387002	equivalent_to	MONDO

---

Output Data

PROJECT_FOLDER
├── input
│   └── ...
└── output
    ├── domain_ontology
    │   ├── edges_hierarchy.csv
    │   ├── mappings.csv
    │   ├── merges.csv
    │   └── nodes.csv
    ├── intermediate
    │   ├── data_tests
    │   │   └── ...
    │   ├── dropped_mappings
    │   │   └── equivalence_1_MONDO.csv
    │   ├── ...
    └── report
        ├── data_docs
        │   └── ...
        ├── data_profile_reports
        │   ├── nodes_report.html
        │   └── ...
        ├── index.html
        └── logs
            └── onto-merger.logger

Folders

• domain_ontology: contains the output of the alignment process, i.e. the

  final merged ontology in table format.
• intermediate: contains the intermediate files generated during the

  alignment process, data testing and data profiling.
• report: contains the report of the alignment process with links to the

  data profiling and data testing pages. Also includes the log output file:

  onto-merger.logger.

Table edges_hierarchy.csv

Same format as in the input set. Contains the merged ontology hierarchy.

Table mappings.csv

Same format as in the input set. Node IDs are potentially updated with canonical node IDs and / or internal code reassignments.

Table merges.csv

source_id	target_id
SNOMED:31387002	MONDO:0004979

Table nodes.csv

Same format as in the input set. Contains the final set of nodes (i.e. canonical nodes). Merged nodes are excluded from this table.