Upload README.md

README.md

@@ -46,31 +46,16 @@ configs:
 
 ## Dataset Description
 
-This dataset is designed for training the Prot2Text framework. It  contains information for 256,690 proteins and is comprised of three modalities:
+This dataset is designed for training the [Prot2Text](https://huggingface.co/habdine/Prot2Text-Base-v1-0) framework. It  contains information for 256,690 proteins and is comprised of three modalities:
 
-Protein sequence (amino acid sequence)
-AlphaFold accession ID
-Textual description of the protein
-The dataset is built from the SwissProt database, a component of UniProtKB Release 2022_04.
+* Protein sequence (amino acid sequence).
+* Protein structure using its AlphaFold accession ID.
+* Textual description of the protein
 
-## Dataset Summary
+The dataset is built from the SwissProt database, a component of UniProtKB Release 2022_04.
 
-Format: TSV (Tab-separated values) - expected (not explicitly mentioned)
-Splits: Train (248,315), Validation (4,172), Test (4,203)
-Size: 256,690 proteins
-Features:
-name (str): Full name of the protein
-sequence (str): Amino acid sequence of the protein
-AlphaFoldDB (str): Accession ID of the protein in AlphaFold database
-taxon (str): Species information (potentially)
-text (str): Textual description of the protein (excluding "(By Similarity)" and "PubMed")
-## Supported Tasks and Leaderboards
 
-This dataset is intended for training and evaluating models that learn relationships between protein sequences, structures, and textual descriptions. Tasks may include:
 
-Protein function prediction
-Protein structure prediction
-Text-based protein information retrieval
 ## Dataset Structure
 
 #### Data Fields:
@@ -91,7 +76,7 @@ The dataset is split into training, validation, and test sets with a maximum seq
 * **Validation:** 4,172 proteins
 * **Test:** 4,203 proteins
 
-## Considerations for Using the Data
+### Considerations for Using the Data
 
 The dataset is built from a single source (SwissProt). Consider incorporating data from other sources to increase diversity.
 The textual descriptions may contain biases present in the original database. Be mindful of these biases when using the data for downstream tasks.
@@ -104,6 +89,13 @@ We are releasing this dataset under the terms of [CC-BY-NC-4.0](https://opendata
 ### Citation
 Please cite this dataset and the original sources if you use this dataset in your work
 ```none
-
+@inproceedings{abdine2024prot2text,
+  title={Prot2Text: Multimodal Protein's Function Generation with GNNs and Transformers},
+  author={Abdine, Hadi and Chatzianastasis, Michail and Bouyioukos, Costas and Vazirgiannis, Michalis},
+  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
+  volume={38},
+  pages={10757--10765},
+  year={2024}
+}
 ```