Hugging Face's logo

Datasets:
Synthyra
/
clustered_ppi_unified

Modalities:
Image
Tags:
protein-protein-interactions
biology
bioinformatics
Dataset card Data Studio Files
xet
Community
Dataset Preview
Duplicate
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 2 new columns ({'value', 'stat'}) and 5 missing columns ({'neg_count', 'organism', 'pos_prop', 'pos_count', 'neg_prop'}).

This happened while the csv dataset builder was generating data using

hf://datasets/Synthyra/clustered_ppi_unified/plots/unified_st030/train_organism_distribution_stats.csv (at revision f1a4d96c14b28e174116e975cc5eb7db425badd4)

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/python3.12/site-packages/datasets/builder.py", line 1831, in _prepare_split_single
                  writer.write_table(table)
                File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 714, in write_table
                  pa_table = table_cast(pa_table, self._schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              stat: string
              value: double
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 478
              to
              {'organism': Value('string'), 'pos_count': Value('int64'), 'neg_count': Value('int64'), 'pos_prop': Value('float64'), 'neg_prop': Value('float64')}
              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 1334, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 911, in stream_convert_to_parquet
                  builder._prepare_split(
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1702, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1833, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
              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 2 new columns ({'value', 'stat'}) and 5 missing columns ({'neg_count', 'organism', 'pos_prop', 'pos_count', 'neg_prop'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/Synthyra/clustered_ppi_unified/plots/unified_st030/train_organism_distribution_stats.csv (at revision f1a4d96c14b28e174116e975cc5eb7db425badd4)
              
              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.

organism
string
pos_count
int64
neg_count
int64
pos_prop
float64
neg_prop
float64
anopheles gambiae
1
0
0
0
apis mellifera
1
0
0
0
arabidopsis thaliana
10,811,460
10,974,212
0.171963
0.174552
bacillus subtilis
15
6
0
0
bos taurus
547
544
0.000009
0.000009
caenorhabditis elegans
7,862,758
7,895,356
0.125062
0.125581
candida albicans
1,183
1,196
0.000019
0.000019
canis familiaris
83
64
0.000001
0.000001
cavia porcellus
4
8
0
0
chlamydomonas reinhardtii
16
14
0
0
cricetulus griseus
41
42
0.000001
0.000001
danio rerio
11,700,048
11,796,854
0.186097
0.187637
dictyostelium discoideum
106
134
0.000002
0.000002
drosophila melanogaster
4,413,456
4,451,704
0.070199
0.070807
emericella nidulans
86
94
0.000001
0.000001
equus caballus
3
6
0
0
escherichia coli
1,167,686
1,081,328
0.018573
0.017199
felis catus
1
0
0
0
gallus gallus
640
698
0.00001
0.000011
glycine max
56
66
0.000001
0.000001
homo sapiens
12,778,277
12,735,256
0.203247
0.202563
human herpesvirus
1,840
1,790
0.000029
0.000028
human immunodeficiency virus
993
108
0.000016
0.000002
human papillomavirus
3,257
2,040
0.000052
0.000032
leishmania major
2
0
0
0
macaca mulatta
286
38
0.000005
0.000001
meleagris gallopavo
3
0
0
0
middle-east respiratory syndrome-related coronavirus
468
64
0.000007
0.000001
monodelphis domestica
1
0
0
0
mus musculus
10,964,740
11,025,930
0.174401
0.175375
mycobacterium tuberculosis
21
20
0
0
neurospora crassa
52
28
0.000001
0
oryctolagus cuniculus
244
224
0.000004
0.000004
oryza sativa
561
652
0.000009
0.00001
pan troglodytes
43
2
0.000001
0
pediculus humanus
3
0
0
0
plasmodium falciparum
3,729
4,038
0.000059
0.000064
rattus norvegicus
13,166
13,936
0.000209
0.000222
ricinus communis
2
0
0
0
saccharomyces cerevisiae
3,005,616
2,765,844
0.047806
0.043993
schizosaccharomyces pombe
110,311
114,158
0.001755
0.001816
selaginella moellendorffii
12
4
0
0
severe acute respiratory syndrome coronavirus 2
23,102
78
0.000367
0.000001
severe acute respiratory syndrome-related coronavirus
1,777
90
0.000028
0.000001
simian immunodeficiency virus
13
6
0
0
simian virus
6
0
0
0
solanum lycopersicum
199
174
0.000003
0.000003
solanum tuberosum
3
0
0
0
sorghum bicolor
2
0
0
0
streptococcus pneumoniae
6
0
0
0
sus scrofa
77
76
0.000001
0.000001
tobacco mosaic virus
1
0
0
0
vaccinia virus
276
282
0.000004
0.000004
vitis vinifera
2
0
0
0
xenopus laevis
3,420
3,532
0.000054
0.000056
zea mays
20
26
0
0
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
End of preview.

Clustered PPI datasets (BIOGRID + STRING) with sequence-disjoint splits

This dataset repo contains multiple dataset variants of protein–protein interactions (PPIs), built by clustering proteins by sequence similarity and then constructing train/valid/test splits that are intended to be disjoint at the protein level (and thus hard to memorize via near-identical sequences).

Artifacts are stored as compressed pickles (*.pkl.gz). A helper downloader exists in this repo:

  • data_processing/download_ppi_data.py::download_clustered_ppi_data

What’s in each split dataframe?

Each split is a pandas.DataFrame with (at minimum):

  • IdA / IdB: protein identifiers
  • OrgA / OrgB: organism identifiers (STRING taxon id for STRING datasets; BIOGRID org id for BIOGRID datasets)
  • labels: >0 indicates a positive interaction, 0 indicates a sampled negative

Some variants also include additional columns (e.g. cluster_a, cluster_b, confidences, org_a, org_b). When negatives are concatenated, some of these columns may be NaN for negative rows.

Dataset variants (index)

A machine-readable index is available at:

  • tables/dataset_index.csv
variant source threshold train rows valid rows test rows train pos rate protein overlap (max)
unified_st030 unified st030 62870722 20124 200056 0.500 0
unified_st040 unified st040 62588678 20020 200000 0.500 0
unified_st050 unified st050 62365306 20024 200240 0.500 0
unified_st060 unified st060 62397798 20094 200044 0.500 0
unified_st070 unified st070 62340120 20004 200012 0.500 0
unified_st080 unified st080 62375686 20014 200000 0.500 0
unified_st090 unified st090 62310384 20010 200042 0.500 0
unified_st095 unified st095 62351836 20008 200104 0.500 0

Per-variant deep dive (plots + stats)

Each variant has:

  • plots/<variant>/...png (rendered below)
  • tables/<variant>/summary.csv and tables/<variant>/schema.csv

unified_st030

Open report

Summary tables

  • tables/unified_st030/summary.csv
  • tables/unified_st030/schema.csv

Label balance

  • train: plots/unified_st030/train_label_counts.png
  • valid: plots/unified_st030/valid_label_counts.png
  • test: plots/unified_st030/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st030/train_organism_distribution.csv
  • stats: plots/unified_st030/train_organism_distribution_stats.csv

  • data: plots/unified_st030/valid_organism_distribution.csv
  • stats: plots/unified_st030/valid_organism_distribution_stats.csv

  • data: plots/unified_st030/test_organism_distribution.csv
  • stats: plots/unified_st030/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st030/cross_split_pos_stats.csv
  • negatives: plots/unified_st030/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st030/train_seq_length_stats.csv

  • stats: plots/unified_st030/valid_seq_length_stats.csv

  • stats: plots/unified_st030/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st030/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st030/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st030/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st030/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st030/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st030/test_top_org_pairs_neg.png

unified_st040

Open report

Summary tables

  • tables/unified_st040/summary.csv
  • tables/unified_st040/schema.csv

Label balance

  • train: plots/unified_st040/train_label_counts.png
  • valid: plots/unified_st040/valid_label_counts.png
  • test: plots/unified_st040/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st040/train_organism_distribution.csv
  • stats: plots/unified_st040/train_organism_distribution_stats.csv

  • data: plots/unified_st040/valid_organism_distribution.csv
  • stats: plots/unified_st040/valid_organism_distribution_stats.csv

  • data: plots/unified_st040/test_organism_distribution.csv
  • stats: plots/unified_st040/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st040/cross_split_pos_stats.csv
  • negatives: plots/unified_st040/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st040/train_seq_length_stats.csv

  • stats: plots/unified_st040/valid_seq_length_stats.csv

  • stats: plots/unified_st040/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st040/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st040/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st040/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st040/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st040/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st040/test_top_org_pairs_neg.png

unified_st050

Open report

Summary tables

  • tables/unified_st050/summary.csv
  • tables/unified_st050/schema.csv

Label balance

  • train: plots/unified_st050/train_label_counts.png
  • valid: plots/unified_st050/valid_label_counts.png
  • test: plots/unified_st050/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st050/train_organism_distribution.csv
  • stats: plots/unified_st050/train_organism_distribution_stats.csv

  • data: plots/unified_st050/valid_organism_distribution.csv
  • stats: plots/unified_st050/valid_organism_distribution_stats.csv

  • data: plots/unified_st050/test_organism_distribution.csv
  • stats: plots/unified_st050/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st050/cross_split_pos_stats.csv
  • negatives: plots/unified_st050/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st050/train_seq_length_stats.csv

  • stats: plots/unified_st050/valid_seq_length_stats.csv

  • stats: plots/unified_st050/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st050/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st050/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st050/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st050/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st050/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st050/test_top_org_pairs_neg.png

unified_st060

Open report

Summary tables

  • tables/unified_st060/summary.csv
  • tables/unified_st060/schema.csv

Label balance

  • train: plots/unified_st060/train_label_counts.png
  • valid: plots/unified_st060/valid_label_counts.png
  • test: plots/unified_st060/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st060/train_organism_distribution.csv
  • stats: plots/unified_st060/train_organism_distribution_stats.csv

  • data: plots/unified_st060/valid_organism_distribution.csv
  • stats: plots/unified_st060/valid_organism_distribution_stats.csv

  • data: plots/unified_st060/test_organism_distribution.csv
  • stats: plots/unified_st060/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st060/cross_split_pos_stats.csv
  • negatives: plots/unified_st060/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st060/train_seq_length_stats.csv

  • stats: plots/unified_st060/valid_seq_length_stats.csv

  • stats: plots/unified_st060/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st060/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st060/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st060/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st060/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st060/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st060/test_top_org_pairs_neg.png

unified_st070

Open report

Summary tables

  • tables/unified_st070/summary.csv
  • tables/unified_st070/schema.csv

Label balance

  • train: plots/unified_st070/train_label_counts.png
  • valid: plots/unified_st070/valid_label_counts.png
  • test: plots/unified_st070/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st070/train_organism_distribution.csv
  • stats: plots/unified_st070/train_organism_distribution_stats.csv

  • data: plots/unified_st070/valid_organism_distribution.csv
  • stats: plots/unified_st070/valid_organism_distribution_stats.csv

  • data: plots/unified_st070/test_organism_distribution.csv
  • stats: plots/unified_st070/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st070/cross_split_pos_stats.csv
  • negatives: plots/unified_st070/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st070/train_seq_length_stats.csv

  • stats: plots/unified_st070/valid_seq_length_stats.csv

  • stats: plots/unified_st070/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st070/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st070/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st070/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st070/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st070/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st070/test_top_org_pairs_neg.png

unified_st080

Open report

Summary tables

  • tables/unified_st080/summary.csv
  • tables/unified_st080/schema.csv

Label balance

  • train: plots/unified_st080/train_label_counts.png
  • valid: plots/unified_st080/valid_label_counts.png
  • test: plots/unified_st080/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st080/train_organism_distribution.csv
  • stats: plots/unified_st080/train_organism_distribution_stats.csv

  • data: plots/unified_st080/valid_organism_distribution.csv
  • stats: plots/unified_st080/valid_organism_distribution_stats.csv

  • data: plots/unified_st080/test_organism_distribution.csv
  • stats: plots/unified_st080/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st080/cross_split_pos_stats.csv
  • negatives: plots/unified_st080/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st080/train_seq_length_stats.csv

  • stats: plots/unified_st080/valid_seq_length_stats.csv

  • stats: plots/unified_st080/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st080/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st080/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st080/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st080/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st080/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st080/test_top_org_pairs_neg.png

unified_st090

Open report

Summary tables

  • tables/unified_st090/summary.csv
  • tables/unified_st090/schema.csv

Label balance

  • train: plots/unified_st090/train_label_counts.png
  • valid: plots/unified_st090/valid_label_counts.png
  • test: plots/unified_st090/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st090/train_organism_distribution.csv
  • stats: plots/unified_st090/train_organism_distribution_stats.csv

  • data: plots/unified_st090/valid_organism_distribution.csv
  • stats: plots/unified_st090/valid_organism_distribution_stats.csv

  • data: plots/unified_st090/test_organism_distribution.csv
  • stats: plots/unified_st090/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st090/cross_split_pos_stats.csv
  • negatives: plots/unified_st090/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st090/train_seq_length_stats.csv

  • stats: plots/unified_st090/valid_seq_length_stats.csv

  • stats: plots/unified_st090/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st090/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st090/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st090/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st090/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st090/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st090/test_top_org_pairs_neg.png

unified_st095

Open report

Summary tables

  • tables/unified_st095/summary.csv
  • tables/unified_st095/schema.csv

Label balance

  • train: plots/unified_st095/train_label_counts.png
  • valid: plots/unified_st095/valid_label_counts.png
  • test: plots/unified_st095/test_label_counts.png

Organism distributions (positives vs negatives)

  • data: plots/unified_st095/train_organism_distribution.csv
  • stats: plots/unified_st095/train_organism_distribution_stats.csv

  • data: plots/unified_st095/valid_organism_distribution.csv
  • stats: plots/unified_st095/valid_organism_distribution_stats.csv

  • data: plots/unified_st095/test_organism_distribution.csv
  • stats: plots/unified_st095/test_organism_distribution_stats.csv

Cross-split organism shift tests

  • positives: plots/unified_st095/cross_split_pos_stats.csv
  • negatives: plots/unified_st095/cross_split_neg_stats.csv

Sequence length distributions (unique proteins)

  • stats: plots/unified_st095/train_seq_length_stats.csv

  • stats: plots/unified_st095/valid_seq_length_stats.csv

  • stats: plots/unified_st095/test_seq_length_stats.csv

Top organism pairs

  • train positives: plots/unified_st095/train_top_org_pairs_pos.png

  • train negatives: plots/unified_st095/train_top_org_pairs_neg.png

  • valid positives: plots/unified_st095/valid_top_org_pairs_pos.png

  • valid negatives: plots/unified_st095/valid_top_org_pairs_neg.png

  • test positives: plots/unified_st095/test_top_org_pairs_pos.png

  • test negatives: plots/unified_st095/test_top_org_pairs_neg.png

How to download and load

Use the helper in this codebase:

from data_processing.download_ppi_data import download_clustered_ppi_data

# BIOGRID example
train_df, valid_df, test_df, interaction_set, seq_dict = download_clustered_ppi_data(
    data_type='biogrid',
    cluster_percentage=0.5,
    hf_repo='Synthyra/clustered_ppi_unified_3',
)

# STRING example (descriptor must match the variant prefix: e.g. 'human' or 'model_orgs')
train_df, valid_df, test_df, interaction_set, seq_dict = download_clustered_ppi_data(
    data_type='string',
    descriptor='human',
    cluster_percentage=0.5,
    hf_repo='Synthyra/clustered_ppi_unified_3',
)
Downloads last month
312