Datasets:

clear-bio
/

ingmar-atlas

Tags:

License:

Dataset card Data Studio Files Files and versions

xet

Community

Dataset Preview

Duplicate

Split (1)

train

The full dataset viewer is not available (click to read why). Only showing a preview of the rows.

The dataset generation failed because of a cast error

Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 5 new columns ({'cut', 'pid', 'stratum', 'protease', 'protease_class'}) and 2 missing columns ({'accession', 'genus'}).

This happened while the csv dataset builder was generating data using

hf://datasets/clear-bio/ingmar-atlas/atlas_v5/protease_class_by_cut_v5.csv (at revision c99a3bcee22e935e514c85e1e24bbc10b9d6cff2), ['hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/newparent_folds_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/protease_class_by_cut_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/protease_site_atlas_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/uniprot_harvest_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/lockbox_candidates_v3.csv']

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1848, in _prepare_split_single
                  writer.write_table(table)
                  ~~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
                  self._write_table(pa_table, writer_batch_size=writer_batch_size)
                  ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              pid: string
              cut: double
              fold: int64
              family: string
              protease: string
              stratum: string
              protease_class: string
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 1068
              to
              {'accession': Value('string'), 'family': Value('string'), 'genus': Value('float64'), 'fold': Value('int64')}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1850, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
                  ...<4 lines>...
                  )
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 5 new columns ({'cut', 'pid', 'stratum', 'protease', 'protease_class'}) and 2 missing columns ({'accession', 'genus'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/clear-bio/ingmar-atlas/atlas_v5/protease_class_by_cut_v5.csv (at revision c99a3bcee22e935e514c85e1e24bbc10b9d6cff2), ['hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/newparent_folds_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/protease_class_by_cut_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/protease_site_atlas_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/atlas_v5/uniprot_harvest_v5.csv', 'hf://datasets/clear-bio/ingmar-atlas@c99a3bcee22e935e514c85e1e24bbc10b9d6cff2/lockbox_candidates_v3.csv']
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

accession string	family string	genus null	fold int64
Q6SW62	Herpesvirales	null	1
P16753	Herpesvirales	null	1
P11306	Retroviridae	null	4
Q0Q4F2	Coronaviridae	null	4
P11268	Retroviridae	null	3
Q0ZME7	Coronaviridae	null	2
Q66814	Filoviridae	null	0
P11261	Retroviridae	null	4
Q66810	Filoviridae	null	2
P15777	Coronaviridae	null	1
Q66799	Filoviridae	null	2
Q66798	Filoviridae	null	0
P29791	Pneumoviridae	null	2
P11225	Coronaviridae	null	4
P16046	Herpesvirales	null	1
P11224	Coronaviridae	null	1
P16082	Retroviridae	null	4
P05135	Coronaviridae	null	0
P15073	Retroviridae	null	1
P11209	Pneumoviridae	null	3
P12568	Pneumoviridae	null	3
P23728	Pneumoviridae	null	2
P08359	Retroviridae	null	3
Q7T9D9	Filoviridae	null	0
P08360	Retroviridae	null	4
P87666	Filoviridae	null	2
Q9QAR5	Coronaviridae	null	1
Q02385	Coronaviridae	null	4
Q0Q475	Coronaviridae	null	2
Q3I5J5	Coronaviridae	null	2
Q27ID8	Retroviridae	null	0
Q2F7J1	Retroviridae	null	0
Q2F7I8	Retroviridae	null	0
P26804	Retroviridae	null	1
P03391	Retroviridae	null	4
P03390	Retroviridae	null	1
P03388	Retroviridae	null	1
P25190	Coronaviridae	null	1
P03386	Retroviridae	null	0
P03420	Pneumoviridae	null	3
P25191	Coronaviridae	null	1
P25192	Coronaviridae	null	1
Q14EB0	Coronaviridae	null	2
P25194	Coronaviridae	null	1
Q3LZX1	Coronaviridae	null	2
P26803	Retroviridae	null	1
Q64756	Adenoviridae	null	1
P12650	Coronaviridae	null	1
P12651	Coronaviridae	null	1
P04027	Retroviridae	null	3
P12722	Coronaviridae	null	0
Q5SC53	Paramyxoviridae	null	0
P13843	Pneumoviridae	null	4
Q05320	Filoviridae	null	2
P06751	Retroviridae	null	4
P06752	Retroviridae	null	2
P36334	Coronaviridae	null	0
P25193	Coronaviridae	null	1
Q8JSP8	Coronaviridae	null	2
O11457	Filoviridae	null	2
K9N5Q8	Coronaviridae	null	4
P32541	Retroviridae	null	4
P33481	Paramyxoviridae	null	0
P21436	Retroviridae	null	1
P21415	Retroviridae	null	3
P22432	Coronaviridae	null	4
Q8BB25	Coronaviridae	null	2
P21445	Retroviridae	null	3
P22429	Retroviridae	null	4
P09458	Paramyxoviridae	null	0
P21443	Retroviridae	null	1
P22428	Retroviridae	null	4
P22427	Retroviridae	null	4
P22167	Pneumoviridae	null	2
Q9JAE0	Paramyxoviridae	null	0
P22430	Retroviridae	null	4
P06445	Retroviridae	null	4
P0DTC2	Coronaviridae	null	0
Q8V436	Coronaviridae	null	1
P51520	Retroviridae	null	4
P31794	Retroviridae	null	3
Q9IKD1	Coronaviridae	null	2
P51515	Retroviridae	null	4
P31791	Retroviridae	null	3
P87671	Filoviridae	null	2
O36634	Pneumoviridae	null	4
Q9QAQ8	Coronaviridae	null	1
P59594	Coronaviridae	null	1
Q91DD8	Filoviridae	null	2
P19716	Paramyxoviridae	null	0
Q89853	Filoviridae	null	2
Q91A26	Coronaviridae	null	1
P26095	Orthomyxoviridae	null	4
P26097	Orthomyxoviridae	null	4
P31626	Retroviridae	null	2
P31627	Retroviridae	null	4
P31789	Retroviridae	null	4
P26094	Orthomyxoviridae	null	4
P26032	Paramyxoviridae	null	0
P33615	Paramyxoviridae	null	1

End of preview.

ingmar-atlas — annotated viral protein cleavage sites

14,314 annotated protease cleavage junctions across viral proteins, harmonized from NCBI RefSeq viral genomes, ViralZone, an 8-family polyprotein corpus, and a UniProt non-polyprotein harvest (v5 extension: +1,778 cuts / +1,174 proteins, adding TTSP-, AVP-, assemblin- and SKI-1-processed substrates), with per-site protease attribution, P4–P4′ octamers, family/genus taxonomy, and cluster-held-out fold assignments. Built for the clear-bio/ingmar cleavage-site predictor; useful on its own for viral protease analyses.

Files

file	rows	content
`atlas_v5/protease_site_atlas_v5.csv`	14,314	one row per annotated junction: source, accession, organism, family, genus, pos_in_parent (0-based P1′), flanking product tags, octamer (P4–P4′), protease label, stratum, parent length
`atlas_v5/protease_class_by_cut_v5.csv`	6,325	dev-corpus cuts joined to chemistry classes (Cys-3C-like, Cys-NIa-Pro, Flavi-Ser, host-signalase, TTSP, AVP, assemblin, …)
`atlas_v5/newparent_folds_v5.csv`	1,174	parent → fold assignment for v5-added parents (cluster-held-out, leakage-free; v4 parents never move folds)
`atlas_v5/uniprot_harvest_v5.csv`	1,778	the v5 UniProt non-polyprotein harvest slice (Signal/Site features → cut conventions documented in `notes/atlas_v5_extension.md`)
`lockbox_candidates_v3.csv`	3,994	held-out lockbox candidate set with exact labels (`label`) and legacy ±2-tolerant labels (`label_pm2`)
`MANIFEST.sha256`	85	integrity hashes for the ingmar data tree

Integrity

Octamers re-derived from parent sequences: 100% exact where parents are available (5,591/5,591 refseq+poly8; 936/936 verifiable viralzone).
Lockbox parents are disjoint from all training folds (verified).
v5 package table exact-deduplicated (4,555 → 4,547 class rows; the v4 table carried 8 fully-identical duplicate rows — documented, negligible effect).
Full audit: notes/integrity_audit.md + notes/integrity_audit_v3.csv (47 checks) in the GitHub repo.

Caveats

22 duplicate (accession, pos, protease) rows are deliberate pp1a/pp1ab judgment-call pairs (same cut annotated on both overlapping ORFs).
unassigned protease labels are honest abstentions, not missing data.
The atlas inherits the taxonomic skew of RefSeq viral reference genomes (picornavirus/potyvirus/coronavirus heavy); the v5 harvest adds Herpesvirales-, Retroviridae-, Orthomyxoviridae-rich glycoprotein maturation cuts.

Provenance

Sources: NCBI RefSeq viral division (10.1093/nar/gkae1038), ViralZone (10.1093/nar/gkq901), UniProt (10.1093/nar/gkae1010). Extraction and harmonization code: clear-bio/ingmar GitHub repo.

Downloads last month: 26

Total file size:

3.49 MB