Regenerate README after dropping pride_accession

README.md

@@ -6,7 +6,7 @@ tags:
   - biology
   - mass-spectrometry
   - proteomics
-  - de-novo-peptide-sequencing
+  - de-novo-sequencing
 size_categories:
   - 1M<n<10M
 configs:
@@ -56,8 +56,6 @@ dataset_info:
       dtype: string
     - name: "localisation_options"
       dtype: int64
-    - name: "pride_accession"
-      dtype: string
   splits:
     - name: train
       num_examples: 6992871
@@ -71,8 +69,9 @@ dataset_info:
 
 ## Overview
 
-Peptide-spectrum match (PSM) data assembled from 18 public PRIDE proteomics projects, partitioned by sequence identity into train, validation, and test splits for de novo peptide sequencing benchmarks. 
-This dataset was part of the [training data](https://instadeepai.github.io/InstaNovo/reference/models/?h=pride#training-datasets) for the InstaNovo model ([Eloff, Kalogeropoulos et al., *Nature Machine Intelligence*, 2025]((https://doi.org/10.1038/s42256-025-01019-5))).
+Peptide-spectrum match (PSM) data assembled from 18 public PRIDE proteomics projects, partitioned by sequence identity into train, validation, and test splits for de novo peptide sequencing benchmarks. **This dataset was used to train [InstaNovo](https://doi.org/10.1038/s42256-025-01019-5)** (Eloff, Kalogeropoulos et al., *Nature Machine Intelligence*, 2025).
+
+Curated by Jeroen Van Goey (j.vangoey@instadeep.com).
 
 ## Dataset Details
 
@@ -107,14 +106,13 @@ This dataset was part of the [training data](https://instadeepai.github.io/Insta
 | `Qvalue` | `double` |
 | `TD` | `large_string` |
 | `Scan number` | `int64` |
-| `mz_array` | `large_list<item: double>` |
-| `intensity_array` | `large_list<item: double>` |
+| `mz_array` | `large_list<element: double>` |
+| `intensity_array` | `large_list<element: double>` |
 | `PXD_identifier` | `large_string` |
 | `modification_string` | `large_string` |
 | `localisation_options` | `int64` |
-| `pride_accession` | `string` |
 
-`pride_accession` is added by the build pipeline from the source filename so each row can be traced back to its PRIDE project.
+Each row carries its source PRIDE accession in the `PXD_identifier` column.
 
 ## Dataset Sources
 
@@ -143,9 +141,7 @@ Per-PRIDE-project breakdown with original publications:
 
 ## Acknowledgements
 
-This HuggingFace dataset was created by Jeroen Van Goey.
-
-Big thanks to Pathmanaban Ramasamy, Tine Claeys, and Lennart Martens of the CompOmics research group for providing us with this training data.
+> Big thanks to Pathmanaban Ramasamy, Tine Claeys, and Lennart Martens of the CompOmics research group for providing us with this training data.
 
 ## Citation