Update README.md

README.md

@@ -88,6 +88,56 @@ This dataset contains curated subsets of various thermal stability measurements
 
 Useful for training models to predict various thermal stability metrics, or evaluating stability effects of mutations.
 
+
+## Quickstart Usage
+### Install HuggingFace Datasets package
+Each subset can be loaded into python using the Huggingface [datasets](https://huggingface.co/docs/datasets/index) library.
+First, from the command line install the `datasets` library
+
+    $ pip install datasets
+
+then, from within python load the datasets library
+
+    >>> import datasets
+
+
+### Load Dataset
+
+Load one of the 'RosettaCommons/FireProtDB2' datasets.
+
+    >>> mutation_dg = datasets.load_dataset('RosettaCommons/FireProtDB2', name = 'mutation_dg')
+    Downloading data: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████| 11.7M/11.7M [00:00<00:00, 13.8MB/s]
+    Downloading data: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████| 1.89M/1.89M [00:00<00:00, 6.71MB/s]
+    Downloading data: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████| 1.88M/1.88M [00:00<00:00, 3.74MB/s]
+    Generating train split: 100%|████████████████████████████████████████████████████████████████████████████| 410591/410591 [00:00<00:00, 1268242.15 examples/s]
+    Generating validation split: 100%|█████████████████████████████████████████████████████████████████████████| 50465/50465 [00:00<00:00, 1280516.59 examples/s]
+    Generating test split: 100%|███████████████████████████████████████████████████████████████████████████████| 50115/50115 [00:00<00:00, 1557480.33 examples/s]
+
+and the dataset is loaded as a `datasets.arrow_dataset.Dataset`
+
+    >>> mutation_dg
+    DatasetDict({
+        train: Dataset({
+            features: ['sequence_length', 'protein_name', 'organism', 'uniprotkb', 'ec_number', 'interpro', 'pmid', 'doi', 'publication_year', 'source_dataset', 'referencing_dataset', 'wwpdb_raw', 'ddg', 'domainome_ddg', 'dg', 'dh', 'dhvh', 'tm', 'dtm', 'exp_temperature', 'fitness', 'ph', 'buffer_norm', 'method_norm', 'measure_norm', 'stabilizing', 'buffer_raw', 'buffer_conc_raw', 'ion_raw', 'ion_conc_raw', 'state', 'pdb_id', 'pdb_ids', 'wt_residue', 'position', 'mut_residue', 'mutation', 'sequence', 'sequence_len_uniprot', 'sequence_length_num', 'length_match', 'protein_id', 'cluster_id', 'split'],
+            num_rows: 410591
+        })
+        validation: Dataset({
+            features: ['sequence_length', 'protein_name', 'organism', 'uniprotkb', 'ec_number', 'interpro', 'pmid', 'doi', 'publication_year', 'source_dataset', 'referencing_dataset', 'wwpdb_raw', 'ddg', 'domainome_ddg', 'dg', 'dh', 'dhvh', 'tm', 'dtm', 'exp_temperature', 'fitness', 'ph', 'buffer_norm', 'method_norm', 'measure_norm', 'stabilizing', 'buffer_raw', 'buffer_conc_raw', 'ion_raw', 'ion_conc_raw', 'state', 'pdb_id', 'pdb_ids', 'wt_residue', 'position', 'mut_residue', 'mutation', 'sequence', 'sequence_len_uniprot', 'sequence_length_num', 'length_match', 'protein_id', 'cluster_id', 'split'],
+            num_rows: 50465
+        })
+        test: Dataset({
+            features: ['sequence_length', 'protein_name', 'organism', 'uniprotkb', 'ec_number', 'interpro', 'pmid', 'doi', 'publication_year', 'source_dataset', 'referencing_dataset', 'wwpdb_raw', 'ddg', 'domainome_ddg', 'dg', 'dh', 'dhvh', 'tm', 'dtm', 'exp_temperature', 'fitness', 'ph', 'buffer_norm', 'method_norm', 'measure_norm', 'stabilizing', 'buffer_raw', 'buffer_conc_raw', 'ion_raw', 'ion_conc_raw', 'state', 'pdb_id', 'pdb_ids', 'wt_residue', 'position', 'mut_residue', 'mutation', 'sequence', 'sequence_len_uniprot', 'sequence_length_num', 'length_match', 'protein_id', 'cluster_id', 'split'],
+            num_rows: 50115
+        })
+    })
+
+which is a column oriented format that can be accessed directly, converted in to a `pandas.DataFrame`, or `parquet` format, e.g.
+    
+    >>> mutation_dg.data.column('protein_name')
+    >>> mutation_dg.to_pandas()
+    >>> mutation_dg.to_parquet("dataset.parquet")
+
+
 ## Dataset Structure
 
 Subsets included are: