Hugging Face's logo

Datasets:
potteryrage
/
SLp-1.1-data

Tasks:
Other
Languages:
English
Size:
1M<n<10M
Tags:
biology
genomics
perturbation
synthetic-lethality
world-model
License:
Dataset card Files
xet
 Community
Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
aggregate_reconstruction_checks: list<item: struct<action_id: string, cells: int64, context_id: string, expected_cells: int64, index: (... 106 chars omitted)
  child 0, item: struct<action_id: string, cells: int64, context_id: string, expected_cells: int64, index: int64, kin (... 94 chars omitted)
      child 0, action_id: string
      child 1, cells: int64
      child 2, context_id: string
      child 3, expected_cells: int64
      child 4, index: int64
      child 5, kind: string
      child 6, protein_max_abs_error: double
      child 7, rna_max_abs_error: double
      child 8, target_guide_set: string
counts: struct<cells: int64, excluded_validation_cells: int64, protein_molecular_channels: int64, reconstruc (... 173 chars omitted)
  child 0, cells: int64
  child 1, excluded_validation_cells: int64
  child 2, protein_molecular_channels: int64
  child 3, reconstruction_train_cells: int64
  child 4, reconstruction_validation_cells: int64
  child 5, rna_nnz: int64
  child 6, rna_queries: int64
  child 7, shards: int64
  child 8, source_train_cells: int64
  child 9, verified_control_cells: int64
identity: string
limitations: list<item: string>
  child 0, item: string
profile: struct<admitted_values_read: int64, columns: int64, elapsed_seconds: double, projected_rna_scan_seco (... 73 chars omitted)
  child 0, admitted_values_read: int64
  child 1, columns: int64
  child 2, elapsed_seconds: double
  child 3, projected_rna_scan_seconds: double
  child 4, source_entries_scanned: 
...
row_stop: int64, row (... 26 chars omitted)
  child 0, item: struct<bytes: int64, path: string, rna_nnz: int64, row_start: int64, row_stop: int64, rows: int64, s (... 14 chars omitted)
      child 0, bytes: int64
      child 1, path: string
      child 2, rna_nnz: int64
      child 3, row_start: int64
      child 4, row_stop: int64
      child 5, rows: int64
      child 6, sha256: string
source_hashes: struct<data/derived/slp11-frangieh/paired-development-v1/adt-channel-roster.json: string, data/deriv (... 381 chars omitted)
  child 0, data/derived/slp11-frangieh/paired-development-v1/adt-channel-roster.json: string
  child 1, data/derived/slp11-frangieh/paired-development-v1/development.npz: string
  child 2, data/derived/slp11-frangieh/paired-development-v1/paired-cell-access.npz: string
  child 3, data/derived/slp11-frangieh/paired-development-v1/rna-query-ensembl-ids.txt: string
  child 4, data/sources/frangieh-2021-scp1064-v1/FrangiehIzar2021_RNA.h5ad: string
  child 5, data/sources/frangieh-2021-scp1064-v1/FrangiehIzar2021_protein.h5ad: string
status: string
channels: list<item: struct<channel_id: string, matched_isotype_label: string, protein_label: string, role: st (... 6 chars omitted)
  child 0, item: struct<channel_id: string, matched_isotype_label: string, protein_label: string, role: string>
      child 0, channel_id: string
      child 1, matched_isotype_label: string
      child 2, protein_label: string
      child 3, role: string
source: string
taxonomy: int64
to
{'channels': List({'channel_id': Value('string'), 'matched_isotype_label': Value('string'), 'protein_label': Value('string'), 'role': Value('string')}), 'identity': Value('string'), 'schema': Value('string'), 'source': Value('string'), 'taxonomy': Value('int64')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_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
              aggregate_reconstruction_checks: list<item: struct<action_id: string, cells: int64, context_id: string, expected_cells: int64, index: (... 106 chars omitted)
                child 0, item: struct<action_id: string, cells: int64, context_id: string, expected_cells: int64, index: int64, kin (... 94 chars omitted)
                    child 0, action_id: string
                    child 1, cells: int64
                    child 2, context_id: string
                    child 3, expected_cells: int64
                    child 4, index: int64
                    child 5, kind: string
                    child 6, protein_max_abs_error: double
                    child 7, rna_max_abs_error: double
                    child 8, target_guide_set: string
              counts: struct<cells: int64, excluded_validation_cells: int64, protein_molecular_channels: int64, reconstruc (... 173 chars omitted)
                child 0, cells: int64
                child 1, excluded_validation_cells: int64
                child 2, protein_molecular_channels: int64
                child 3, reconstruction_train_cells: int64
                child 4, reconstruction_validation_cells: int64
                child 5, rna_nnz: int64
                child 6, rna_queries: int64
                child 7, shards: int64
                child 8, source_train_cells: int64
                child 9, verified_control_cells: int64
              identity: string
              limitations: list<item: string>
                child 0, item: string
              profile: struct<admitted_values_read: int64, columns: int64, elapsed_seconds: double, projected_rna_scan_seco (... 73 chars omitted)
                child 0, admitted_values_read: int64
                child 1, columns: int64
                child 2, elapsed_seconds: double
                child 3, projected_rna_scan_seconds: double
                child 4, source_entries_scanned: 
              ...
              row_stop: int64, row (... 26 chars omitted)
                child 0, item: struct<bytes: int64, path: string, rna_nnz: int64, row_start: int64, row_stop: int64, rows: int64, s (... 14 chars omitted)
                    child 0, bytes: int64
                    child 1, path: string
                    child 2, rna_nnz: int64
                    child 3, row_start: int64
                    child 4, row_stop: int64
                    child 5, rows: int64
                    child 6, sha256: string
              source_hashes: struct<data/derived/slp11-frangieh/paired-development-v1/adt-channel-roster.json: string, data/deriv (... 381 chars omitted)
                child 0, data/derived/slp11-frangieh/paired-development-v1/adt-channel-roster.json: string
                child 1, data/derived/slp11-frangieh/paired-development-v1/development.npz: string
                child 2, data/derived/slp11-frangieh/paired-development-v1/paired-cell-access.npz: string
                child 3, data/derived/slp11-frangieh/paired-development-v1/rna-query-ensembl-ids.txt: string
                child 4, data/sources/frangieh-2021-scp1064-v1/FrangiehIzar2021_RNA.h5ad: string
                child 5, data/sources/frangieh-2021-scp1064-v1/FrangiehIzar2021_protein.h5ad: string
              status: string
              channels: list<item: struct<channel_id: string, matched_isotype_label: string, protein_label: string, role: st (... 6 chars omitted)
                child 0, item: struct<channel_id: string, matched_isotype_label: string, protein_label: string, role: string>
                    child 0, channel_id: string
                    child 1, matched_isotype_label: string
                    child 2, protein_label: string
                    child 3, role: string
              source: string
              taxonomy: int64
              to
              {'channels': List({'channel_id': Value('string'), 'matched_isotype_label': Value('string'), 'protein_label': Value('string'), 'role': Value('string')}), 'identity': Value('string'), 'schema': Value('string'), 'source': Value('string'), 'taxonomy': Value('int64')}
              because column names don't match

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.

SLp-1.1 prepared research data

Version r1 contains the prepared inputs used to train SLp-1.1's molecular and functional world components, plus separate downstream-decoder inputs and evaluation evidence. Model · Code and reproduction · Scientific model card.

The file hierarchy mirrors the source checkout. Download with python scripts/fetch_artifacts.py data from a cloned source repository. artifacts.lock.json pins this release to an immutable HF commit. The helper verifies SHA-256 checksums and refuses to overwrite changed local files. inventory.json is the complete file list, with byte sizes, roles and source terms. Model inference only requires the separate model download.

Group Contents
Molecular fitting data/derived/slp11-cell-world-training-v5/ and all referenced fitting arrays: sparse paired cells, memory-mapped RNA counts and population responses
Functional fitting/development data/derived/slp11-genomic-fitness-world-v1/: 4,182,121 observed human fitting effects across 843 contexts and 1,818,947 yeast fitting pairs; development observations are stored separately
Application benchmark Selected MuSL folds, roster and descriptor inputs; these labels are for the separate downstream decoder, never world pretraining
Evidence Fixed contexts, per-fold decoder outputs/controls and molecular/functional evaluation reports
Rights Original source receipts and additional source-scope notices

Human identifiers retain NCBI taxonomy 9606; yeast identifiers retain 4932. Human cell RNA is processed log1p(CP10K), not raw sequencing reads. Yeast RNA targets are population means of per-cell log1p(CP10K). Human fitness targets are raw DepMap gene effects. Yeast binary records retain relative single/double fitness; the trainer applies its recorded floor and natural logarithm. Other population assays retain the units in their manifests.

This is a prepared-input release, not a mirror of every upstream raw dataset or every historical SLp experiment. The molecular index enforces its own fitting rows even when a source file also contains held rows. Development and benchmark files must not be added to fitting. Historical local paths in receipts describe where preparation occurred; executable loaders use explicit repository-relative input directories. The development guide gives the supported training commands.

See THIRD_PARTY_NOTICES.md for source-specific licenses, citations, adaptations and exclusions. MIT applies to original SLp code and weights, not to these third-party biological observations. The GEO components retain the NCBI public-data-policy designation and submitter rights.
Downloads last month
99
Total file size:
11.2 GB

Models trained or fine-tuned on potteryrage/SLp-1.1-data