Datasets:

DOI:
License:
ChrisMungall commited on
Commit
f7fe9a6
·
verified ·
1 Parent(s): 7352796

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +51 -3
README.md CHANGED
@@ -1,3 +1,51 @@
1
- ---
2
- license: cc0-1.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc0-1.0
3
+ ---
4
+ # CHEBI Chemical Classification Program Ontology (C3PO) Benchmark
5
+
6
+ Name: C3PO-v237
7
+
8
+ A benchmark for chemical classification, derived from v237 of the CHEBI ontology.
9
+
10
+ The benchmark consists of
11
+
12
+ - 177,875 *Structures*, each represented by a SMILES string
13
+ - 1364 *Classes*, each of which classify (directly or indirectly) between 25 and 5000 Structures
14
+
15
+ The CHEBI smiles property (obo:chebi/smiles) is used to derive
16
+ structures. Only SMILES strings that lacked a `*` wildcard are
17
+ used. If multiple CHEBI entries have the same SMILES, they are
18
+ collapsed.
19
+
20
+ The benchmark can be further partitioned based on different properties.
21
+
22
+ C3PO-Slim is the subset of structures that have at least one xref
23
+ (mapping), representing the set of classes that are more commonly used
24
+ in the biological literature and in other databases.
25
+
26
+ 20% of the structures are designated as being for *validation*
27
+ purposes. The remaining 80% can be used for training and testing.
28
+
29
+ The benchmark can be downloaded as either a single JSON file, or
30
+ separate CSVs for classes and structures. Summary stats files are also
31
+ provided:
32
+
33
+ - [dataset.json](dataset.json) - the complete benchmark
34
+ - [classes.csv](classes.csv) - all classes
35
+ - [classes_slim.csv](classes_slim.csv) - slim classes (any with an xref)
36
+ - [structures.csv](structures.csv) - all structures
37
+ - [classes_summary.csv](classes_summary.csv) - summary statistics for classes
38
+ - [classes_slim_summary.csv](classes_slim_summary.csv) - summary statistics for classes
39
+ - [structures_summary.csv](structures_summary.csv) - summary statistics for structures
40
+ - [metadata.csv](metadata.csv) - metadata about how this was derived.
41
+
42
+ For a description of all fields, see:
43
+
44
+ - [datamodel](https://github.com/cmungall/c3p/blob/main/c3p/datamodel.py)
45
+
46
+ For the code that extracts the benchmark from CHEBI, see:
47
+
48
+ - [extractor](https://github.com/cmungall/c3p/blob/main/c3p/extractor.py)
49
+
50
+
51
+