Upload README.md
Browse files
README.md
CHANGED
|
@@ -8,8 +8,6 @@ tags:
|
|
| 8 |
- classification
|
| 9 |
size_categories:
|
| 10 |
- 10K<n<100K
|
| 11 |
-
task_categories:
|
| 12 |
-
- sequence-classification
|
| 13 |
license: cc-by-4.0
|
| 14 |
---
|
| 15 |
|
|
@@ -28,7 +26,7 @@ A curated collection of vertebrate UniProt/Swiss‑Prot protein *domains* labele
|
|
| 28 |
- **Splits:** Stratified train/test/val = 70%/20%/10% on the representative set.
|
| 29 |
- **Total size (full set):** **10,652** domains (4,635 cytosolic only; 5,917 extracellular only; 100 with both labels).
|
| 30 |
|
| 31 |
-
|
| 32 |
|
| 33 |
### Supported Tasks
|
| 34 |
- **Sequence classification:** Predicting whether a protein domain is viable in the cytosol/extracellular space.
|
|
@@ -45,7 +43,8 @@ A curated collection of vertebrate UniProt/Swiss‑Prot protein *domains* labele
|
|
| 45 |
- `full_dataset.csv` — Full curated set (**10,652** domains).
|
| 46 |
- `representative_data_train.csv`, `representative_data_val.csv`, `representative_data_test.csv` — 30%‑identity clustered representative set with predefined splits for model development.
|
| 47 |
|
| 48 |
-
|
|
|
|
| 49 |
|
| 50 |
### Fields
|
| 51 |
- `sequence` *(str)* — Amino‑acid sequence of the domain.
|
|
@@ -53,9 +52,11 @@ A curated collection of vertebrate UniProt/Swiss‑Prot protein *domains* labele
|
|
| 53 |
- `cytosolic_label` *(bool)* — `True` if the domain localizes to the intracellular/cytosolic region.
|
| 54 |
- `extracellular_label` *(bool)* — `True` if the domain localizes to the extracellular space (secreted or extracellular domain of a TM protein).
|
| 55 |
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
|
|
|
|
|
|
| 59 |
|
| 60 |
---
|
| 61 |
|
|
|
|
| 8 |
- classification
|
| 9 |
size_categories:
|
| 10 |
- 10K<n<100K
|
|
|
|
|
|
|
| 11 |
license: cc-by-4.0
|
| 12 |
---
|
| 13 |
|
|
|
|
| 26 |
- **Splits:** Stratified train/test/val = 70%/20%/10% on the representative set.
|
| 27 |
- **Total size (full set):** **10,652** domains (4,635 cytosolic only; 5,917 extracellular only; 100 with both labels).
|
| 28 |
|
| 29 |
+
See the **Files** section for exact filenames and formats.
|
| 30 |
|
| 31 |
### Supported Tasks
|
| 32 |
- **Sequence classification:** Predicting whether a protein domain is viable in the cytosol/extracellular space.
|
|
|
|
| 43 |
- `full_dataset.csv` — Full curated set (**10,652** domains).
|
| 44 |
- `representative_data_train.csv`, `representative_data_val.csv`, `representative_data_test.csv` — 30%‑identity clustered representative set with predefined splits for model development.
|
| 45 |
|
| 46 |
+
|
| 47 |
+
Column schema is consistent across files.
|
| 48 |
|
| 49 |
### Fields
|
| 50 |
- `sequence` *(str)* — Amino‑acid sequence of the domain.
|
|
|
|
| 52 |
- `cytosolic_label` *(bool)* — `True` if the domain localizes to the intracellular/cytosolic region.
|
| 53 |
- `extracellular_label` *(bool)* — `True` if the domain localizes to the extracellular space (secreted or extracellular domain of a TM protein).
|
| 54 |
|
| 55 |
+
Notes:
|
| 56 |
+
|
| 57 |
+
- The two labels are **not mutually exclusive**; a small number of domains have both annotations.
|
| 58 |
+
|
| 59 |
+
- For TM proteins we include **only the annotated extracellular domain** within the length range.
|
| 60 |
|
| 61 |
---
|
| 62 |
|