Update README.md

README.md

@@ -42,44 +42,90 @@ citation_bibtex: "@article{KoehlerLeman2023,\n  title = {Sequence-structure-func
   \  Ian and Xavier,  Ramnik J. and Knight,  Rob and Bonneau,  Richard and Kosciolek,\
   \  Tomasz},\n  year = {2023},\n  month = apr\n}"
 citation_apa: Koehler Leman, J., Szczerbiak, P., Renfrew, P. D., Gligorijevic, V.,
-  Berenberg, D., Vatanen, T., … Kosciolek, T. (2023). Sequence-structure-function
-  relationships in the microbial protein universe. Nature Communications, 14(1), 2351.
-  doi:10.1038/s41467-023-37896-w
-dataset_info:
-  features:
-  - name: id
-    dtype: string
-  - name: pdb
-    dtype: large_string
-  - name: Filter_Stage2_aBefore
-    dtype: float64
-  - name: Filter_Stage2_bQuarter
-    dtype: float64
-  - name: Filter_Stage2_cHalf
-    dtype: float64
-  - name: Filter_Stage2_dEnd
-    dtype: float64
-  - name: clashes_bb
-    dtype: float64
-  - name: clashes_total
-    dtype: float64
-  - name: score
-    dtype: float64
-  - name: silent_score
-    dtype: float64
-  - name: time
-    dtype: float64
-  splits:
-  - name: train
-    num_bytes: 26606148779
-    num_examples: 211069
-  download_size: 9111920009
-  dataset_size: 26606148779
+  Berenberg, D., Vatanen, T., Taylor, B. C., Janssen, S., Pataki, A., Carriero, N.,
+  Fisk, I., Xavier, R. J., Knight, R., Bonneau, R., Kosciolek, T. (2023).
+  Sequence-structure-function relationships in the microbial protein universe.
+  Nature Communications, 14(1), 2351. doi:10.1038/s41467-023-37896-w
+config_names:
+  - rosetta_models
+  - dmpfold_models
+  - rosetta_function_predictions
+  - dmpfold_function_predictions
 configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
+  - config_name: rosetta_models
+    data_files:
+    - split: high_quality
+      path: data/rosetta_high_quality_models-*
+    - split: low_quality
+      path: data/rosetta_low_quality_models-*
+  - config_name: dmpfold_models
+    data_files:
+    - split: high_quality
+      path: data/dmpfold_high_quality_models-*
+    - split: low_quality
+      path: data/dmpfold_low_quality_models-*
+  - config_name: DeepFRI_function_predictions
+    data_files:
+    - split: rosetta_high_quality
+      path: data/rosetta_high_quality_function_predictions-*
+    - split: rosetta_low_quality
+      path: data/rosetta_low_quality_function_predictions-*
+    - split: dmpfold_high_quality
+      path: data/dmpfold_high_quality_function_predictions-*
+    - split: dmpfold_low_quality
+      path: data/dmpfold_low_quality_function_predicitons-*
+dataset_info:
+  - config_name: rosetta_models
+    features:
+      - name: id
+        dtype: string
+      - name: pdb
+        dtype: large_string
+      - name: Filter_Stage2_aBefore
+        dtype: float64
+      - name: Filter_Stage2_bQuarter
+        dtype: fload64
+      - name: Filter_Stage2_cHalf
+        dtype: fload64
+      - name: Filter_Stage2_dEnd
+        dtype: float64
+      - name: clashes_bb
+        dtype: float64
+      - name: clashes_total
+        dtype: float64
+      - name: score
+        dtype: float64
+      - name: silent_score
+        dtype: float64
+      - name: time
+        dtype: float64
+    splits:
+      - name: high_quality
+        num_examples: 23164
+      - name: low_quality
+        num_examples: 309363
+  - config_name: dmpfold_models
+    features:
+      - name: id
+        dtype: string
+      - name: pdb
+        dtype: string
+    split:
+      - name: high_quality
+        num_examples: 203878
+      - name: low_quality
+        num_examples: 37957
+  - config_name: DeepFRI_function_predictions
+    features:
+      - name: id
+        dtype: string
+    splits:
+      - name: rosetta_high_quality
+      - name: rosetta_low_quality
+      - name: dmpfold_high_quality
+      - name: dmpfold_low_quality
+  
+    
 ---
 # Microbiome Immunity Project: Protein Universe
 ~200,000 predicted structures for diverse protein sequences from 1,003
@@ -121,10 +167,7 @@ embeddings from DeepFRI.
 
 - **License:** cc-by-4.0
 
-### Dataset Sources [optional]
-
-<!-- Provide the basic links for the dataset. -->
-
+### Dataset Sources
 - **Repository:** https://github.com/microbiome-immunity-project/protein_universe
 - **Paper:**
   Koehler Leman, J., Szczerbiak, P., Renfrew, P. D., Gligorijevic, V., Berenberg,
@@ -142,19 +185,17 @@ proteins from Archaea and Bacteria, whose protein sequences are generally shorte
 than Eukaryotic.
 
 ### Direct Use
-This dataset could be used to
+This dataset could be used to train representation models of protein structure
 
 -
 
 
 ### Out-of-Scope Use
-
 While this dataset has been curated for quality, in some cases the predicted structures
 may not represent physically realistic conformations. Thus caution much be used when using
 it as training data for protein structure prediction and design.
 
 ## Dataset Structure
-
     microbiome_immunity_project_dataset
       dataset
         dmpfold_high_quality_function_predictions