diff --git a/src/protify/data/supported_datasets.py b/src/protify/data/supported_datasets.py index 12ec3e1..ccc28ea 100644 --- a/src/protify/data/supported_datasets.py +++ b/src/protify/data/supported_datasets.py @@ -1,4 +1,3 @@ - supported_datasets = { 'EC': 'GleghornLab/EC_reg', 'GO-CC': 'GleghornLab/CC_reg', @@ -6,46 +5,48 @@ 'GO-MF': 'GleghornLab/MF_reg', 'MB': 'GleghornLab/MB_reg', 'DeepLoc-2': 'GleghornLab/DL2_reg', - 'DeepLoc-10': 'GleghornLab/DL10_reg', - 'Subcellular': 'GleghornLab/SL_13', + 'DeepLoc-10': 'GleghornLab/DL10_reg', + 'SL13': 'GleghornLab/SL_13', 'enzyme-kcat': 'GleghornLab/enzyme_kcat', 'solubility': 'GleghornLab/solubility_prediction', - 'localization': 'GleghornLab/localization_prediction', - 'temperature-stability': 'GleghornLab/temperature_stability', - 'peptide-HLA-MHC-affinity': 'GleghornLab/peptide_HLA_MHC_affinity_ppi', - 'optimal-temperature': 'GleghornLab/optimal_temperature', + 'temp-stability': 'GleghornLab/temperature_stability', + 'optimal-temp': 'GleghornLab/optimal_temperature', 'optimal-ph': 'GleghornLab/optimal_ph', - 'optimal-ph-rigor': 'GleghornLab/optimal_ph_rigor', - 'material-production': 'GleghornLab/material_production', - 'fitness-prediction': 'GleghornLab/fitness_prediction', + 'mat-production': 'GleghornLab/material_production', + 'fitness-pred': 'GleghornLab/fitness_prediction', 'number-of-folds': 'GleghornLab/fold_prediction', 'cloning-clf': 'GleghornLab/cloning_clf', - 'stability-prediction': 'GleghornLab/stability_prediction', - 'human-ppi-saprot': 'GleghornLab/HPPI', - 'SecondaryStructure-3': 'GleghornLab/SS3', - 'SecondaryStructure-8': 'GleghornLab/SS8', - 'fluorescence-prediction': 'GleghornLab/fluorescence_prediction', - 'plastic': 'GleghornLab/plastic_degradation_benchmark', - 'gold-ppi': 'Synthyra/bernett_gold_ppi', - 'human-ppi-pinui': 'GleghornLab/HPPI_PiNUI', - 'yeast-ppi-pinui': 'GleghornLab/YPPI_PiNUI', - 'shs27-ppi-raw': 'Synthyra/SHS27k', - 'shs148-ppi-raw': 'Synthyra/SHS148k', - 'shs27-ppi-random': 'GleghornLab/ppi_SHS27k_random_2025', - 'shs148-ppi-random': 'GleghornLab/ppi_SHS148k_random_2025', - 'shs27-ppi-dfs': 'GleghornLab/ppi_SHS27k_dfs_2025', - 'shs148-ppi-dfs': 'GleghornLab/ppi_SHS148k_dfs_2025', - 'shs27-ppi-bfs': 'GleghornLab/ppi_SHS27k_bfs_2025', + 'stability-pred': 'GleghornLab/stability_prediction', + 'ec-active': 'lhallee/ec_active', + 'ecoli_expression': 'GleghornLab/ecoli_expression', + 'soluprot': 'GleghornLab/soluprot', + 'KSMoFinder-clustered': 'GleghornLab/ksmo_clustered', + 'KSMoFinder': 'GleghornLab/KSmo_fixed', 'shs148-ppi-bfs': 'GleghornLab/ppi_SHS148k_bfs_2025', - 'string-ppi-random': 'GleghornLab/ppi_STRING_random_2025', - 'string-ppi-dfs': 'GleghornLab/ppi_STRING_dfs_2025', + 'shs27-ppi-bfs': 'GleghornLab/ppi_SHS27k_bfs_2025', 'string-ppi-bfs': 'GleghornLab/ppi_STRING_bfs_2025', + 'gold-ppi': 'Synthyra/bernett_gold_ppi', 'plm-interact': 'GleghornLab/plm_interact_human_train_cross_ppi', - 'ppi-mutation-effect': 'GleghornLab/ppi_mutation_effect', # requires multi_column 'PPA-ppi': 'Synthyra/ppi_affinity', + 'million_full': 'GleghornLab/millionfull_round_1_oct_2025', + 'taxon_species': 'GleghornLab/taxonomy_species_0.4_clusters', + 'diff_phylogeny': 'GleghornLab/diff_phylo', + 'localization': 'GleghornLab/localization_prediction', + 'peptide-HLA-MHC-affinity': 'GleghornLab/peptide_HLA_MHC_affinity_ppi', + 'optimal-ph-rigor': 'GleghornLab/optimal_ph_rigor', + 'SS3': 'GleghornLab/SS3', + 'SS8': 'GleghornLab/SS8', 'foldseek-fold': 'lhallee/foldseek_dataset', # prostt5 'foldseek-inverse': 'lhallee/foldseek_dataset', # prostt5 - 'ec-active': 'lhallee/ec_active', + 'plddt': 'GleghornLab/af2_plddt', + 'shs27-ppi-random': 'GleghornLab/ppi_SHS27k_random_2025', + 'shs27-ppi-dfs': 'GleghornLab/ppi_SHS27k_dfs_2025', + 'shs27-ppi-raw': 'Synthyra/SHS27k', + 'shs148-ppi-random': 'GleghornLab/ppi_SHS148k_random_2025', + 'shs148-ppi-dfs': 'GleghornLab/ppi_SHS148k_dfs_2025', + 'shs148-ppi-raw': 'Synthyra/SHS148k', + 'string-ppi-random': 'GleghornLab/ppi_STRING_random_2025', + 'string-ppi-dfs': 'GleghornLab/ppi_STRING_dfs_2025', 'proteingym_zs': 'proteingym_zs', # not a path, data loading for this is currently handled in benchmarks/proteingym/data_loader.py 'proteingym_supervised': 'proteingym_supervised', # not a path, data loading for this is currently handled in benchmarks/proteingym/data_loader.py 'taxon_domain': 'GleghornLab/taxonomy_domain_0.4_clusters', @@ -55,15 +56,12 @@ 'taxon_order': 'GleghornLab/taxonomy_order_0.4_clusters', 'taxon_family': 'GleghornLab/taxonomy_family_0.4_clusters', 'taxon_genus': 'GleghornLab/taxonomy_genus_0.4_clusters', - 'taxon_species': 'GleghornLab/taxonomy_species_0.4_clusters', - 'diff_phylogeny': 'GleghornLab/diff_phylo', - 'plddt': 'GleghornLab/af2_plddt', + 'human-ppi-saprot': 'GleghornLab/HPPI', + 'human-ppi-pinui': 'GleghornLab/HPPI_PiNUI', + 'yeast-ppi-pinui': 'GleghornLab/YPPI_PiNUI', + 'ppi-mutation-effect': 'GleghornLab/ppi_mutation_effect', # requires multi_column + 'fluorescence': 'GleghornLab/fluorescence_prediction', 'realness': 'GleghornLab/realness_dataset', - 'million_full': 'GleghornLab/millionfull_round_1_oct_2025', - 'soluprot': 'GleghornLab/soluprot', - 'ecoli_expression': 'GleghornLab/ecoli_expression', - 'KSMoFinder-clustered': 'GleghornLab/ksmo_clustered', - 'KSMoFinder': 'GleghornLab/KSmo_fixed', # bom-pooling paper (Hoang & Singh 2025) datasets 'FLUO': 'GleghornLab/bom_fluorescence', 'BLAC': 'GleghornLab/bom_blac', @@ -71,11 +69,6 @@ 'DPI': 'GleghornLab/bom_dpi', } -internal_datasets = { - 'plastic': 'GleghornLab/plastic_degradation_benchmark', -} - -# TODO update possible_with_vector_reps = [ ### multi-label 'EC', @@ -83,31 +76,44 @@ 'GO-CC', 'GO-BP', 'GO-MF', - 'Subcellular', - # ppi - 'shs27-ppi-random', - 'shs27-ppi-dfs', - 'shs27-ppi-bfs', - 'shs148-ppi-random', - 'shs148-ppi-dfs', - 'shs148-ppi-bfs', - 'string-ppi-random', - 'string-ppi-dfs', - 'string-ppi-bfs', + 'SL13', ### classification 'MB', 'DeepLoc-2', 'DeepLoc-10', 'solubility', - 'temperature-stability', - 'material-production', - 'fitness-prediction', + 'temp-stability', + 'mat-production', + 'fitness-pred', 'number-of-folds', 'cloning-clf', - 'stability-prediction', + 'stability-pred', 'ec-active', 'localization', - # taxonomy + ### regression + 'enzyme-kcat', + 'optimal-temp', + 'optimal-ph', + 'million_full', + ### ppi + 'shs148-ppi-bfs', + 'shs27-ppi-bfs', + 'string-ppi-bfs', + 'gold-ppi', + 'plm-interact', + 'PPA-ppi', + 'shs27-ppi-raw', + 'shs148-ppi-raw', + 'shs27-ppi-random', + 'shs27-ppi-dfs', + 'shs148-ppi-random', + 'shs148-ppi-dfs', + 'string-ppi-random', + 'string-ppi-dfs', + 'human-ppi-saprot', + 'human-ppi-pinui', + 'yeast-ppi-pinui', + ### taxonomy 'taxon_domain', 'taxon_kingdom', 'taxon_phylum', @@ -117,22 +123,6 @@ 'taxon_genus', 'taxon_species', 'diff_phylogeny', - # ppi - 'shs27-ppi-raw', - 'shs148-ppi-raw', - 'plm-interact', - 'gold-ppi', - 'string-ppi-bfs', - 'human-ppi-saprot', - 'human-ppi-pinui', - 'yeast-ppi-pinui', - ### regression - 'enzyme-kcat', - 'optimal-temperature', - 'optimal-ph', - 'million_full', - # ppi - 'PPA-ppi', ] # TODO update @@ -146,9 +136,9 @@ 'DeepLoc-2', 'DeepLoc-10', 'enzyme-kcat', - 'optimal-temperature', + 'opt-temp', 'optimal-ph', - 'fitness-prediction', + 'fitness-pred', ] @@ -159,37 +149,35 @@ 'GO-CC', # SaProt 'GO-BP', # SaProt 'GO-MF', # SaProt - 'Subcellular', # Custom - Tamar - # ppi - 'shs27-ppi-bfs', # MGPPI + SHS27k + SHS148k + STRING - 'shs148-ppi-bfs', # MGPPI + SHS27k + SHS148k + STRING - 'string-ppi-bfs', # MGPPI + SHS27k + SHS148k + STRING + 'SL13', # Custom - Tamar ### classification 'MB', # SaProt 'DeepLoc-2', # SaProt 'DeepLoc-10', # SaProt 'solubility', # Biomap - 'temperature-stability', # Biomap - 'material-production', # Biomap - 'fitness-prediction', # Biomap + 'temp-stability', # Biomap + 'mat-production', # Biomap + 'fitness-pred', # Biomap 'number-of-folds', # Biomap 'cloning-clf', # Biomap - 'stability-prediction', # Biomap + 'stability-pred', # Biomap 'ec-active', # Custom - Logan 'soluprot', # Custom - SoluProt - # taxonomy + ### taxonomy 'taxon_species', # Accidental taxonomists 'diff_phylogeny', # Accidental taxonomists - # ppi + ### ppi + 'shs27-ppi-bfs', # MGPPI + SHS27k + SHS148k + STRING + 'shs148-ppi-bfs', # MGPPI + SHS27k + SHS148k + STRING + 'string-ppi-bfs', # MGPPI + SHS27k + SHS148k + STRING 'plm-interact', # PLM-Interact 'gold-ppi', # Bernett - ### regression + 'PPA-ppi', # Custom - Logan - Bindwell + # regression 'enzyme-kcat', # Custom - Logan - Biomap - 'optimal-temperature', # Biomap + 'opt-temp', # Biomap 'optimal-ph', # Biomap 'million_full', # Millionfull - # ppi - 'PPA-ppi', # Custom - Logan - Bindwell ] @@ -200,5 +188,5 @@ 'enzyme-kcat', # regression 'human-ppi', # ppi 'plddt', # tokenwise regression - 'SecondaryStructure-3', # tokenwise classification -] \ No newline at end of file + 'SS3', # tokenwise classification +]