Update README.md
Browse files
README.md
CHANGED
@@ -1,39 +1,38 @@
|
|
1 |
-
---
|
2 |
-
license: mit
|
3 |
-
---
|
4 |
-
# Description
|
5 |
-
Binary Localization prediction is a binary classification task where each input protein *x* is mapped to a label *y* ∈ {0, 1}, corresponding to either "membrane-bound" or "soluble" .
|
6 |
-
|
7 |
-
The digital label means:
|
8 |
-
|
9 |
-
0: membrane-bound
|
10 |
-
|
11 |
-
1: soluble
|
12 |
-
|
13 |
-
# Splits
|
14 |
-
|
15 |
-
**Structure type:** AF2
|
16 |
-
|
17 |
-
The dataset is from [**DeepLoc: prediction of protein subcellular localization using deep learning**](https://academic.oup.com/bioinformatics/article/33/21/3387/3931857). We employ all proteins (proteins that lack AF2 structures are removed), and split them based on 70% structure similarity (see [ProteinShake](https://github.com/BorgwardtLab/proteinshake/tree/main)), with the number of training, validation and test set shown below:
|
18 |
-
|
19 |
-
- Train: 6707
|
20 |
-
- Valid: 698
|
21 |
-
- Test: 807
|
22 |
-
|
23 |
-
# Data format
|
24 |
-
|
25 |
-
We organize all data in LMDB format. The architecture of the databse is like:
|
26 |
-
|
27 |
-
**length:** The number of samples
|
28 |
-
|
29 |
-
**0:**
|
30 |
-
|
31 |
-
- **name:** The UniProt ID of the protein
|
32 |
-
|
33 |
-
- **seq:** The structure-aware sequence
|
34 |
-
- **
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
**···**
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
---
|
4 |
+
# Description
|
5 |
+
Binary Localization prediction is a binary classification task where each input protein *x* is mapped to a label *y* ∈ {0, 1}, corresponding to either "membrane-bound" or "soluble" .
|
6 |
+
|
7 |
+
The digital label means:
|
8 |
+
|
9 |
+
0: membrane-bound
|
10 |
+
|
11 |
+
1: soluble
|
12 |
+
|
13 |
+
# Splits
|
14 |
+
|
15 |
+
**Structure type:** AF2
|
16 |
+
|
17 |
+
The dataset is from [**DeepLoc: prediction of protein subcellular localization using deep learning**](https://academic.oup.com/bioinformatics/article/33/21/3387/3931857). We employ all proteins (proteins that lack AF2 structures are removed), and split them based on 70% structure similarity (see [ProteinShake](https://github.com/BorgwardtLab/proteinshake/tree/main)), with the number of training, validation and test set shown below:
|
18 |
+
|
19 |
+
- Train: 6707
|
20 |
+
- Valid: 698
|
21 |
+
- Test: 807
|
22 |
+
|
23 |
+
# Data format
|
24 |
+
|
25 |
+
We organize all data in LMDB format. The architecture of the databse is like:
|
26 |
+
|
27 |
+
**length:** The number of samples
|
28 |
+
|
29 |
+
**0:**
|
30 |
+
|
31 |
+
- **name:** The UniProt ID of the protein
|
32 |
+
|
33 |
+
- **seq:** The structure-aware sequence
|
34 |
+
- **label:** classification label of the sequence
|
35 |
+
|
36 |
+
**1:**
|
37 |
+
|
|
|
38 |
**···**
|