Datasets:

bilalsm commited on
Commit
c646e2e
·
verified ·
1 Parent(s): 9e2276d

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +3 -1
README.md CHANGED
@@ -18,10 +18,12 @@ Converting irregularly spaced mass spectrometry (MS/MS) data into computable rep
18
 
19
  ## Dataset
20
 
 
 
21
  - `inference_data/Massive-KB-v1_test_sorted.h5`: 107,713 held-out spectra.
22
  - `inference_data/PXD017407_library_from_fdr_peptides_sorted.h5`: 2,301 additional spectra.
23
 
24
- This repository contains inference data only. Each H5 file contains `peaks` (`[N, 200, 2]` m/z and intensity), `sequence`, `charge`, `precursor_mass` (precursor m/z), and `protein`. Peptide sequences can include the lowercase modification codes described in the code repository.
25
 
26
  ## Download
27
 
 
18
 
19
  ## Dataset
20
 
21
+ - `training_data/`: 30 MassiveKB H5 files, 1,938,843 spectra.
22
+ - `val_data/Massive-KB-v1_val.h5`: 107,713 validation spectra.
23
  - `inference_data/Massive-KB-v1_test_sorted.h5`: 107,713 held-out spectra.
24
  - `inference_data/PXD017407_library_from_fdr_peptides_sorted.h5`: 2,301 additional spectra.
25
 
26
+ Each H5 file contains `peaks` (`[N, 200, 2]` m/z and intensity), `sequence`, `charge`, `precursor_mass` (precursor m/z), and `protein`. Peptide sequences can include the lowercase modification codes described in the code repository.
27
 
28
  ## Download
29