neuralninja110's picture
Super-squash branch 'main' using huggingface_hub
6174986 verified
|
Raw
History Blame
1.52 kB
---
license: mit
task_categories:
- token-classification
language:
- en
tags:
- protein
- bioinformatics
- secondary-structure
- deep-learning
size_categories:
- 1K<n<10K
---
# Protein Secondary Structure Prediction Dataset (NPPE2)
## Dataset Description
This dataset contains protein sequences with their corresponding secondary structure labels for both Q8 (8-class) and Q3 (3-class) classification tasks.
### Dataset Statistics
| Split | Sequences |
|-------|-----------|
| Train | 7262 |
| Test | 1816 |
### Features
- **id**: Unique identifier for each protein sequence
- **seq**: Amino acid sequence (20 standard amino acids)
- **sst8**: 8-class secondary structure labels (Q8)
- **sst3**: 3-class secondary structure labels (Q3)
### Secondary Structure Labels
**Q8 Labels (DSSP):**
- H = Alpha helix
- E = Beta strand
- C = Coil/Loop
- G = 3-10 helix
- B = Beta bridge
- T = Turn
- S = Bend
- I = Pi helix
**Q3 Labels:**
- H = Helix (H, G, I)
- E = Strand (E, B)
- C = Coil (C, T, S)
### Usage
`python
from datasets import load_dataset
dataset = load_dataset("neuralninja110/dlgenai-nppe-dataset")
`
### Files
- train.csv: Training data with sequences and labels
- test.csv: Test data (sequences only)
- sample_submission.csv: Submission format template
- data/train.parquet: Training data in parquet format
### Evaluation Metric
Harmonic mean of F1 scores for Q8 and Q3 predictions.
### License
MIT License