starkAhmed43 commited on
Commit
e25b3ec
·
verified ·
1 Parent(s): c9260a0

Leak-CURBER leaderboard split: train+val release

Browse files
Files changed (2) hide show
  1. README.md +86 -0
  2. data/train.parquet +3 -0
README.md ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-nc-sa-4.0
3
+ pretty_name: Leak-CURBER Km P-struct
4
+ tags:
5
+ - biology
6
+ - enzymes
7
+ - chemistry
8
+ - leak-curber
9
+ - benchmark
10
+ size_categories:
11
+ - 10K<n<100K
12
+ configs:
13
+ - config_name: default
14
+ data_files:
15
+ - split: train
16
+ path: data/train.parquet
17
+ ---
18
+
19
+ # Leak-CURBER Km P-struct
20
+
21
+ ## Summary
22
+
23
+ Leak-CURBER is a benchmark for enzymatic reaction tasks (NeurIPS 2026). This repository holds the leaderboard split Km P-struct at threshold 0.69. It contains the paper's train and validation rows, merged and shuffled with seed 42. The test set is held out for the leaderboard.
24
+
25
+ ## Rows
26
+
27
+ 74,904 rows in one `train` split.
28
+
29
+ ## Split
30
+
31
+ The P-struct split at threshold 0.69: test proteins have a maximum structural similarity to the training proteins below 0.69 (Foldseek).
32
+
33
+ ## Columns
34
+
35
+ | Column | Type | Meaning |
36
+ |---|---|---|
37
+ | `ECNumber` | large_string | EC number of the measurement (null for some rows). |
38
+ | `substrate` | large_string | Substrate name. |
39
+ | `value` | double | Raw value in the unit given in `unit`. |
40
+ | `Organism` | large_string | Source organism. |
41
+ | `uniprot_id` | large_string | UniProt accession of the enzyme. |
42
+ | `value_type` | large_string | Measured quantity (constant in this file). |
43
+ | `unit` | large_string | Unit of `value` (constant in this file). |
44
+ | `enzymeType` | large_string | wildtype or mutant. |
45
+ | `mutations` | large_list<element: large_string> | List of mutations (null for wild type). |
46
+ | `substrate_id` | large_string | Compound id in the source database (null for some rows). |
47
+ | `uniprot_match` | large_string | How the UniProt id was matched (helper column). |
48
+ | `source` | large_string | Source database. Merged records use `a|b`. |
49
+ | `pdbs` | large_list<element: large_string> | List of PDB ids (empty or null for many rows). |
50
+ | `sequence` | large_string | Amino acid sequence of the enzyme (protein input). |
51
+ | `smiles` | large_string | SMILES of the substrate (molecule input). |
52
+ | `log10_value` | double | Target: log10 of `value`. |
53
+ | `smiles_key` | large_string | Canonical SMILES used for grouping (helper column). |
54
+ | `smiles_hash` | large_string | SHA-256 of `smiles_key` (helper column). |
55
+ | `uniprot_date` | timestamp[ns] | UniProt entry creation date (helper column; used by the Time split). |
56
+
57
+ ## Target
58
+
59
+ `log10_value`, the log10 of the Michaelis constant in mM.
60
+
61
+ ## Sources and license
62
+
63
+ This dataset is released for academic, non-commercial use under CC BY-NC-SA 4.0. The terms of each source database also apply. Users must follow them.
64
+
65
+ - BRENDA, CC BY 4.0 (https://www.brenda-enzymes.org/license.php).
66
+ - SABIO-RK, HITS terms for internal non-commercial research and academic use (https://sabiork.h-its.org/layouts/content/termscondition.gsp). This source restricts use to academic, non-commercial research.
67
+ - UniProt, CC BY 4.0 (https://www.uniprot.org/help/license): protein sequences and annotation dates.
68
+ - RCSB PDB, CC0 1.0 (https://www.rcsb.org/pages/usage-policy), and AlphaFold DB, CC BY 4.0 (https://alphafold.ebi.ac.uk/faq): structure ids in the `pdbs` column refer to these databases, where present.
69
+
70
+ The `source` column gives the database of each row. A value such as `brenda|sabio` means the merged row comes from both. Rows per `source` value: brenda 51,616; sabio 21,382; brenda|sabio 1,906.
71
+
72
+ ## Citation
73
+
74
+ ```bibtex
75
+ @inproceedings{ahmed2026leakcurber,
76
+ title = {Leak-CURBER: A Leakage-Controlled Multimodal Evaluation Benchmark for Enzymatic Reaction Tasks},
77
+ author = {Ahmed, Adhil and Balaji, Roshan and Bhatt, Nirav Pravinbhai},
78
+ booktitle = {Advances in Neural Information Processing Systems},
79
+ year = {2026}
80
+ }
81
+ ```
82
+
83
+ Code: https://github.com/bisect-group/Leak-CURBER
84
+
85
+ Please also cite the source databases: BRENDA (the latest BRENDA paper); SABIO-RK; UniProt (The UniProt Consortium, "UniProt: the Universal Protein Knowledgebase", current Nucleic Acids Research paper); RCSB PDB and the PDB entry authors; AlphaFold DB (Jumper et al., Nature 2021; Varadi et al., Nucleic Acids Research 2024).
86
+
data/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a5d9ea06715614a7b77e75d9a054590836b52012597e722ee98ea3418f4fa85
3
+ size 20725783