Datasets:
Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cdla-permissive-2.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- tabular-classification
|
| 5 |
+
pretty_name: avida_htnfa_onehot
|
| 6 |
+
tags:
|
| 7 |
+
- tabular
|
| 8 |
+
- proteomics
|
| 9 |
+
- binary
|
| 10 |
+
- Biomedical
|
| 11 |
+
- Benchmark
|
| 12 |
+
- antigenomics
|
| 13 |
+
configs:
|
| 14 |
+
- config_name: avida_htnfa_onehot
|
| 15 |
+
data_files:
|
| 16 |
+
- split: train
|
| 17 |
+
path: data/train.parquet
|
| 18 |
+
- split: validation
|
| 19 |
+
path: data/val.parquet
|
| 20 |
+
- split: test
|
| 21 |
+
path: data/test.parquet
|
| 22 |
+
dataset_info:
|
| 23 |
+
description: 'Binary classification: predict VHH nanobody binding to hTNFa. Single
|
| 24 |
+
antigen — antigen features omitted. Features: padded integer AA token IDs of VHH
|
| 25 |
+
sequence (vocab size 21, 0=PAD). Split: CD-HIT cluster-based 80/10/10 on VHH sequences
|
| 26 |
+
(c=0.7, aS=0.8).'
|
| 27 |
+
---
|