From 19942a96dbea5ce7558395bc62e18eed899479e7 Mon Sep 17 00:00:00 2001 From: Satchel Baldwin Date: Wed, 30 Oct 2024 14:25:40 -0500 Subject: [PATCH 1/2] base pdc graphql schema --- .../api_documentation/pdc_schema.graphql | 2728 +++++++++++++++++ .../beaker_biome/integrations/pdc_regex.py | 30 + 2 files changed, 2758 insertions(+) create mode 100644 chat-ui/beaker-biome/src/beaker_biome/biome/api_documentation/pdc_schema.graphql create mode 100644 chat-ui/beaker-biome/src/beaker_biome/integrations/pdc_regex.py diff --git a/chat-ui/beaker-biome/src/beaker_biome/biome/api_documentation/pdc_schema.graphql b/chat-ui/beaker-biome/src/beaker_biome/biome/api_documentation/pdc_schema.graphql new file mode 100644 index 0000000..b045c12 --- /dev/null +++ b/chat-ui/beaker-biome/src/beaker_biome/biome/api_documentation/pdc_schema.graphql @@ -0,0 +1,2728 @@ +query FilteredFilesData( + $program_name_filter: String! + $project_name_filter: String! + $study_name_filter: String! + $disease_filter: String! + $filterValue: String! + $analytical_frac_filter: String! + $exp_type_filter: String! + ) { + uiFile( + program_name: $program_name_filter + project_name: $project_name_filter + study_name: $study_name_filter + disease_type: $disease_filter + primary_site: $filterValue + analytical_fraction: $analytical_frac_filter + experiment_type: $exp_type_filter + ) { + submitter_id_name + file_name + study_run_metadata_submitter_id + project_name + file_type + file_size + } + } + +query FilteredFilesDataPaginated( + $study_version_value: String! + $offset_value: Int + $limit_value: Int + $sort_value: String + $program_name_filter: String! + $project_name_filter: String! + $study_name_filter: String! + $disease_filter: String! + $filterValue: String! + $analytical_frac_filter: String! + $exp_type_filter: String! + $ethnicity_filter: String! + $race_filter: String! + $gender_filter: String! + $tumor_grade_filter: String! + $sample_type_filter: String! + $acquisition_type_filter: String! + $data_category_filter: String! + $file_type_filter: String! + $access_filter: String! + $downloadable_filter: String!, + $biospecimen_status_filter: String!, + $case_status_filter: String! + ) { + getPaginatedUIFile( + study_version: $study_version_value + offset: $offset_value + limit: $limit_value + sort: $sort_value + program_name: $program_name_filter + project_name: $project_name_filter + study_name: $study_name_filter + disease_type: $disease_filter + primary_site: $filterValue + analytical_fraction: $analytical_frac_filter + experiment_type: $exp_type_filter + ethnicity: $ethnicity_filter + race: $race_filter + gender: $gender_filter + tumor_grade: $tumor_grade_filter + sample_type: $sample_type_filter + acquisition_type: $acquisition_type_filter + data_category: $data_category_filter + file_type: $file_type_filter + access: $access_filter + downloadable: $downloadable_filter, + biospecimen_status: $biospecimen_status_filter, + case_status: $case_status_filter + ) { + total + uiFiles { + submitter_id_name + study_id + pdc_study_id + embargo_date + file_name + study_run_metadata_submitter_id + project_name + data_category + file_type + access + file_size + file_id + md5sum + downloadable + } + pagination { + count + sort + from + page + total + pages + size + } + } + } + +query FilteredFilesLegacyDataPaginated( + $offset_value: Int + $limit_value: Int + $sort_value: String! + $study_id_filter: String! + $data_source_filter: String! + $data_category_filter: String! + $file_type_filter: String!) { + getPaginatedUILegacyFile( + study_id: $study_id_filter + data_source: $data_source_filter + offset: $offset_value + limit: $limit_value + sort: $sort_value + data_category:$data_category_filter + file_type:$file_type_filter + ) { + total + uiLegacyFiles { + file_id + pdc_study_id + submitter_id_name + embargo_date + file_name + study_run_metadata_submitter_id + project_name + data_category + file_type + file_size + md5sum + downloadable + access + study_id + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query FilesDataQuery( + $file_name: String! + $study_id: String! + ) { + uiFilesPerStudy (file_name: $file_name, study_id: $study_id) { + file_id + file_name + signedUrl { + url + } + } + } + +query LegacyFilesDataQuery( + $file_name: String! + ) { + uiLegacyFilesPerStudy (file_name: $file_name) { + file_id + file_name + signedUrl { + url + } + } + } + +query PaginatedFilesForPublication( + $offset_value: Int + $limit_value: Int + $publication_id: String! + ) { + getPaginatedUIPancancerFiles( + offset: $offset_value + limit: $limit_value + publication_id: $publication_id + ) { + total + uiPancancerFiles { + file_id + file_name + description + characterization + cohorts + related_publications + related_studies + downloadable + } + pagination { + count + sort + from + page + total + pages + size + } + } + } + +query FilesData { + uiFile { + submitter_id_name + file_name + study_run_metadata_submitter_id + project_name + file_type + file_size + } + } + +query CasePerFile($file_id: String!) { + uiCasePerFile(file_id: $file_id) { + file_id + case_id + case_submitter_id + } + } + +query StudiesVersions{ + getPaginatedUIStudy(offset: 0, limit: 1000){ + uiStudies { + study_id + submitter_id_name + pdc_study_id + versions { + number + } + } + } + } + +query CaseSummaryDataQuery($case_id: String!, $case_submitter_id: String!, $source: String!){ + uiCase (case_id: $case_id, case_submitter_id: $case_submitter_id, source: $source) { + aliquot_id + sample_id + case_id + project_name + program_name + sample_type + disease_type + primary_site + } + } +query FilteredStudiesData($case_id: String!, $source: String!){ + uiCaseSummary(case_id: $case_id, source: $source) { + case_id + case_submitter_id + project_submitter_id + disease_type + tissue_source_site_code + days_to_lost_to_followup + index_date + lost_to_followup + primary_site + demographics { + ethnicity + gender + demographic_submitter_id + race + cause_of_death + days_to_birth + days_to_death + vital_status + year_of_birth + year_of_death + age_at_index + premature_at_birth + weeks_gestation_at_birth + age_is_obfuscated + cause_of_death_source + occupation_duration_years + country_of_residence_at_enrollment + } + samples{ + sample_id + gdc_sample_id + gdc_project_id + sample_submitter_id + sample_type + status + pool + sample_is_ref + biospecimen_anatomic_site + biospecimen_laterality + composition + current_weight + days_to_collection + days_to_sample_procurement + diagnosis_pathologically_confirmed + freezing_method + initial_weight + intermediate_dimension + longest_dimension + method_of_sample_procurement + pathology_report_uuid + preservation_method + sample_type_id + shortest_dimension + time_between_clamping_and_freezing + time_between_excision_and_freezing + tissue_type + tumor_code + tumor_code_id + tumor_descriptor + annotation + aliquots{ + aliquot_id + aliquot_submitter_id + aliquot_quantity + aliquot_volume + status + pool + aliquot_is_ref + amount + analyte_type + concentration + } + + } + } + } +query FilteredStudiesData($case_id: String!, $source: String!){ + uiCaseSummary(case_id: $case_id, source: $source) { + case_id + case_submitter_id + project_submitter_id + disease_type + tissue_source_site_code + days_to_lost_to_followup + index_date + lost_to_followup + primary_site + exposures { + age_at_onset + alcohol_days_per_week + alcohol_drinks_per_day + alcohol_history + alcohol_intensity + alcohol_type + asbestos_exposure + cigarettes_per_day + coal_dust_exposure + environmental_tobacco_smoke_exposure + exposure_id + exposure_submitter_id + exposure_duration + exposure_duration_years + exposure_type + marijuana_use_per_week + pack_years_smoked + parent_with_radiation_exposure + radon_exposure + respirable_crystalline_silica_exposure + secondhand_smoke_as_child + smokeless_tobacco_quit_age + smoking_frequency + time_between_waking_and_first_smoke + tobacco_smoking_onset_year + tobacco_smoking_quit_year + tobacco_smoking_status + tobacco_use_per_day + type_of_smoke_exposure + type_of_tobacco_used + years_smoked + } + follow_ups { + adverse_event + adverse_event_grade + aids_risk_factors + barretts_esophagus_goblet_cells_present + bmi + body_surface_area + cause_of_response + cd4_count + cdc_hiv_risk_factors + comorbidity + comorbidity_method_of_diagnosis + days_to_adverse_event + days_to_comorbidity + days_to_follow_up + days_to_imaging + days_to_progression + days_to_progression_free + days_to_recurrence + diabetes_treatment_type + disease_response + dlco_ref_predictive_percent + ecog_performance_status + evidence_of_recurrence_type + eye_color + fev1_ref_post_bronch_percent + fev1_ref_pre_bronch_percent + fev1_fvc_pre_bronch_percent + fev1_fvc_post_bronch_percent + follow_up_id + follow_up_submitter_id + haart_treatment_indicator + height + hepatitis_sustained_virological_response + history_of_tumor + history_of_tumor_type + hiv_viral_load + hormonal_contraceptive_type + hormonal_contraceptive_use + hormone_replacement_therapy_type + hpv_positive_type + hysterectomy_margins_involved + hysterectomy_type + imaging_result + imaging_type + immunosuppressive_treatment_type + karnofsky_performance_status + menopause_status + nadir_cd4_count + pancreatitis_onset_year + pregnancy_outcome + procedures_performed + progression_or_recurrence + progression_or_recurrence_anatomic_site + progression_or_recurrence_type + recist_targeted_regions_number + recist_targeted_regions_sum + reflux_treatment_type + risk_factor + risk_factor_treatment + scan_tracer_used + undescended_testis_corrected + undescended_testis_corrected_age + undescended_testis_corrected_laterality + undescended_testis_corrected_method + undescended_testis_history + undescended_testis_history_laterality + viral_hepatitis_serologies + weight + } + } + } +query FilteredStudiesData($case_id: String!, $source: String!){ + uiCaseSummary(case_id: $case_id, source: $source) { + case_id + case_submitter_id + project_submitter_id + disease_type + tissue_source_site_code + days_to_lost_to_followup + index_date + lost_to_followup + primary_site + diagnoses{ + age_at_diagnosis + morphology + primary_diagnosis + tumor_grade + tumor_stage + tissue_or_organ_of_origin + diagnosis_submitter_id + classification_of_tumor + days_to_last_follow_up + days_to_last_known_disease_status + days_to_recurrence + last_known_disease_status + progression_or_recurrence + site_of_resection_or_biopsy + prior_malignancy + ajcc_clinical_m + ajcc_clinical_n + ajcc_clinical_stage + ajcc_clinical_t + ajcc_pathologic_m + ajcc_pathologic_n + ajcc_pathologic_stage + ajcc_pathologic_t + ajcc_staging_system_edition + ann_arbor_b_symptoms + ann_arbor_clinical_stage + ann_arbor_extranodal_involvement + ann_arbor_pathologic_stage + best_overall_response + burkitt_lymphoma_clinical_variant + circumferential_resection_margin + colon_polyps_history + days_to_best_overall_response + days_to_diagnosis + days_to_hiv_diagnosis + days_to_new_event + figo_stage + hiv_positive + hpv_positive_type + hpv_status + iss_stage + laterality + ldh_level_at_diagnosis + ldh_normal_range_upper + lymph_nodes_positive + lymphatic_invasion_present + method_of_diagnosis + new_event_anatomic_site + new_event_type + overall_survival + perineural_invasion_present + prior_treatment + progression_free_survival + progression_free_survival_event + residual_disease + vascular_invasion_present + year_of_diagnosis + icd_10_code + synchronous_malignancy + anaplasia_present + anaplasia_present_type + child_pugh_classification + cog_liver_stage + cog_neuroblastoma_risk_group + cog_renal_stage + cog_rhabdomyosarcoma_risk_group + enneking_msts_grade + enneking_msts_metastasis + enneking_msts_stage + enneking_msts_tumor_site + esophageal_columnar_dysplasia_degree + esophageal_columnar_metaplasia_present + first_symptom_prior_to_diagnosis + gastric_esophageal_junction_involvement + goblet_cells_columnar_mucosa_present + gross_tumor_weight + inpc_grade + inpc_histologic_group + inrg_stage + inss_stage + irs_group + irs_stage + ishak_fibrosis_score + lymph_nodes_tested + medulloblastoma_molecular_classification + metastasis_at_diagnosis + metastasis_at_diagnosis_site + mitosis_karyorrhexis_index + peripancreatic_lymph_nodes_positive + peripancreatic_lymph_nodes_tested + supratentorial_localization + tumor_confined_to_organ_of_origin + tumor_focality + tumor_regression_grade + vascular_invasion_type + wilms_tumor_histologic_subtype + breslow_thickness + gleason_grade_group + igcccg_stage + international_prognostic_index + largest_extrapelvic_peritoneal_focus + masaoka_stage + non_nodal_regional_disease + non_nodal_tumor_deposits + ovarian_specimen_status + ovarian_surface_involvement + percent_tumor_invasion + peritoneal_fluid_cytological_status + primary_gleason_grade + secondary_gleason_grade + weiss_assessment_score + adrenal_hormone + ann_arbor_b_symptoms_described + diagnosis_is_primary_disease + eln_risk_classification + figo_staging_edition_year + gleason_grade_tertiary + gleason_patterns_percent + margin_distance + margins_involved_site + pregnant_at_diagnosis + satellite_nodule_present + sites_of_involvement + tumor_depth + who_cns_grade + who_nte_grade + diagnosis_uuid + } + } + } +query ExperimentFileByCaseCountQuery($case_id_filter: String, $source: String!){ + uiExperimentFileCount(case_id: $case_id_filter, source: $source) { + acquisition_type + submitter_id_name + experiment_type + files_count + } + } +query DataCategoryFileByCaseCountQuery($case_id_filter: String, $source: String!){ + uiDataCategoryFileCount (case_id: $case_id_filter, source: $source) { + file_type + submitter_id_name + data_category + files_count + } + } +query FilteredStudiesDataPaginated($study_name_filter: String!, $source: String!){ + getPaginatedUIStudy(study_name: $study_name_filter, source: $source) { + total + uiStudies { + study_id + pdc_study_id + submitter_id_name + study_description + study_submitter_id + program_name + project_name + disease_type + primary_site + analytical_fraction + experiment_type + embargo_date + cases_count + aliquots_count + filesCount { + file_type + data_category + files_count + } + supplementaryFilesCount { + data_category + file_type + files_count + } + nonSupplementaryFilesCount { + data_category + file_type + files_count + } + contacts { + name + institution + email + url + } + versions { + number + } + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query FilteredStudiesData($case_id: String!, $source: String!){ + uiCaseSummary(case_id: $case_id, source: $source) { + case_id + case_submitter_id + project_submitter_id + disease_type + tissue_source_site_code + days_to_lost_to_followup + index_date + lost_to_followup + primary_site + externalReferences { + reference_resource_shortname + reference_entity_location + reference_resource_name + external_reference_id + } + demographics { + ethnicity + gender + demographic_submitter_id + race + cause_of_death + days_to_birth + days_to_death + vital_status + year_of_birth + year_of_death + age_at_index + premature_at_birth + weeks_gestation_at_birth + age_is_obfuscated + cause_of_death_source + occupation_duration_years + country_of_residence_at_enrollment + } + diagnoses{ + samples { + sample_id + sample_submitter_id + annotation + } + age_at_diagnosis + morphology + primary_diagnosis + tumor_grade + tumor_stage + tissue_or_organ_of_origin + diagnosis_submitter_id + classification_of_tumor + days_to_last_follow_up + days_to_last_known_disease_status + days_to_recurrence + last_known_disease_status + progression_or_recurrence + site_of_resection_or_biopsy + prior_malignancy + ajcc_clinical_m + ajcc_clinical_n + ajcc_clinical_stage + ajcc_clinical_t + ajcc_pathologic_m + ajcc_pathologic_n + ajcc_pathologic_stage + ajcc_pathologic_t + ajcc_staging_system_edition + ann_arbor_b_symptoms + ann_arbor_clinical_stage + ann_arbor_extranodal_involvement + ann_arbor_pathologic_stage + best_overall_response + burkitt_lymphoma_clinical_variant + circumferential_resection_margin + colon_polyps_history + days_to_best_overall_response + days_to_diagnosis + days_to_hiv_diagnosis + days_to_new_event + figo_stage + hiv_positive + hpv_positive_type + hpv_status + iss_stage + laterality + ldh_level_at_diagnosis + ldh_normal_range_upper + lymph_nodes_positive + lymphatic_invasion_present + method_of_diagnosis + new_event_anatomic_site + new_event_type + overall_survival + perineural_invasion_present + prior_treatment + progression_free_survival + progression_free_survival_event + residual_disease + vascular_invasion_present + year_of_diagnosis + icd_10_code + synchronous_malignancy + anaplasia_present + anaplasia_present_type + child_pugh_classification + cog_liver_stage + cog_neuroblastoma_risk_group + cog_renal_stage + cog_rhabdomyosarcoma_risk_group + enneking_msts_grade + enneking_msts_metastasis + enneking_msts_stage + enneking_msts_tumor_site + esophageal_columnar_dysplasia_degree + esophageal_columnar_metaplasia_present + first_symptom_prior_to_diagnosis + gastric_esophageal_junction_involvement + goblet_cells_columnar_mucosa_present + gross_tumor_weight + inpc_grade + inpc_histologic_group + inrg_stage + inss_stage + irs_group + irs_stage + ishak_fibrosis_score + lymph_nodes_tested + medulloblastoma_molecular_classification + metastasis_at_diagnosis + metastasis_at_diagnosis_site + mitosis_karyorrhexis_index + peripancreatic_lymph_nodes_positive + peripancreatic_lymph_nodes_tested + supratentorial_localization + tumor_confined_to_organ_of_origin + tumor_focality + tumor_regression_grade + vascular_invasion_type + wilms_tumor_histologic_subtype + breslow_thickness + gleason_grade_group + igcccg_stage + international_prognostic_index + largest_extrapelvic_peritoneal_focus + masaoka_stage + non_nodal_regional_disease + non_nodal_tumor_deposits + ovarian_specimen_status + ovarian_surface_involvement + percent_tumor_invasion + peritoneal_fluid_cytological_status + primary_gleason_grade + secondary_gleason_grade + weiss_assessment_score + adrenal_hormone + ann_arbor_b_symptoms_described + diagnosis_is_primary_disease + eln_risk_classification + figo_staging_edition_year + gleason_grade_tertiary + gleason_patterns_percent + margin_distance + margins_involved_site + pregnant_at_diagnosis + satellite_nodule_present + sites_of_involvement + tumor_depth + who_cns_grade + who_nte_grade + diagnosis_uuid + } + samples{ + sample_id + gdc_sample_id + gdc_project_id + sample_submitter_id + sample_type + status + pool + sample_is_ref + biospecimen_anatomic_site + biospecimen_laterality + composition + current_weight + days_to_collection + days_to_sample_procurement + diagnosis_pathologically_confirmed + freezing_method + initial_weight + intermediate_dimension + longest_dimension + method_of_sample_procurement + pathology_report_uuid + preservation_method + sample_type_id + shortest_dimension + time_between_clamping_and_freezing + time_between_excision_and_freezing + tissue_type + tumor_code + tumor_code_id + tumor_descriptor + annotation + aliquots{ + aliquot_id + aliquot_submitter_id + aliquot_quantity + aliquot_volume + status + pool + aliquot_is_ref + amount + analyte_type + concentration + } + diagnoses { + diagnosis_id + diagnosis_submitter_id + } + } + exposures { + age_at_onset + alcohol_days_per_week + alcohol_drinks_per_day + alcohol_history + alcohol_intensity + alcohol_type + asbestos_exposure + cigarettes_per_day + coal_dust_exposure + environmental_tobacco_smoke_exposure + exposure_id + exposure_submitter_id + exposure_duration + exposure_duration_years + exposure_type + marijuana_use_per_week + pack_years_smoked + parent_with_radiation_exposure + radon_exposure + respirable_crystalline_silica_exposure + secondhand_smoke_as_child + smokeless_tobacco_quit_age + smoking_frequency + time_between_waking_and_first_smoke + tobacco_smoking_onset_year + tobacco_smoking_quit_year + tobacco_smoking_status + tobacco_use_per_day + type_of_smoke_exposure + type_of_tobacco_used + years_smoked + } + follow_ups { + adverse_event + adverse_event_grade + aids_risk_factors + barretts_esophagus_goblet_cells_present + bmi + body_surface_area + cause_of_response + cd4_count + cdc_hiv_risk_factors + comorbidity + comorbidity_method_of_diagnosis + days_to_adverse_event + days_to_comorbidity + days_to_follow_up + days_to_imaging + days_to_progression + days_to_progression_free + days_to_recurrence + diabetes_treatment_type + disease_response + dlco_ref_predictive_percent + ecog_performance_status + evidence_of_recurrence_type + eye_color + fev1_ref_post_bronch_percent + fev1_ref_pre_bronch_percent + fev1_fvc_pre_bronch_percent + fev1_fvc_post_bronch_percent + follow_up_id + follow_up_submitter_id + haart_treatment_indicator + height + hepatitis_sustained_virological_response + history_of_tumor + history_of_tumor_type + hiv_viral_load + hormonal_contraceptive_type + hormonal_contraceptive_use + hormone_replacement_therapy_type + hpv_positive_type + hysterectomy_margins_involved + hysterectomy_type + imaging_result + imaging_type + immunosuppressive_treatment_type + karnofsky_performance_status + menopause_status + nadir_cd4_count + pancreatitis_onset_year + pregnancy_outcome + procedures_performed + progression_or_recurrence + progression_or_recurrence_anatomic_site + progression_or_recurrence_type + recist_targeted_regions_number + recist_targeted_regions_sum + reflux_treatment_type + risk_factor + risk_factor_treatment + scan_tracer_used + undescended_testis_corrected + undescended_testis_corrected_age + undescended_testis_corrected_laterality + undescended_testis_corrected_method + undescended_testis_history + undescended_testis_history_laterality + viral_hepatitis_serologies + weight + } + treatments { + days_to_treatment_end + days_to_treatment_start + initial_disease_status + regimen_or_line_of_therapy + therapeutic_agents + treatment_anatomic_site + treatment_effect + treatment_intent_type + treatment_or_therapy + treatment_outcome + treatment_type + chemo_concurrent_to_radiation + number_of_cycles + reason_treatment_ended + route_of_administration + treatment_arm + treatment_dose + treatment_dose_units + treatment_effect_indicator + treatment_frequency + } + + } + } +query FilteredStudiesData($program_name_filter: String!, $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, $analytical_frac_filter: String!, $exp_type_filter: String!){ + uiStudy(program_name: $program_name_filter , project_name: $project_name_filter, study_name: $study_name_filter, disease_type: $disease_filter, primary_site: $filterValue, analytical_fraction: $analytical_frac_filter, experiment_type: $exp_type_filter) { + submitter_id_name + program_name + project_name + disease_type + primary_site + analytical_fraction + experiment_type + num_raw + num_mzml + num_prot + num_prot_assem + num_psm + } + } +query paginatedUIStudyQuery($study_name: String!, $pdc_study_id: String!, $study_version: String!, $source: String!){ + getPaginatedUIStudy(study_name: $study_name, pdc_study_id: $pdc_study_id, study_version: $study_version, source: $source) { + total + uiStudies { + study_id + pdc_study_id + submitter_id_name + study_description + program_name + project_name + disease_type + primary_site + analytical_fraction + experiment_type + embargo_date + aliquots_count + cases_count + filesCount { + file_type + data_category + files_count + } + supplementaryFilesCount { + data_category + file_type + files_count + } + nonSupplementaryFilesCount { + data_category + file_type + files_count + } + contacts { + name + institution + email + url + } + versions { + number + } + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query WorkflowMetadataQery($study_id: String!, $source: String!){ + uiWorkflowMetadata(study_id: $study_id, source: $source) { + workflow_metadata_submitter_id + study_submitter_id + protocol_submitter_id + cptac_study_id + submitter_id_name + study_submitter_name + analytical_fraction + experiment_type + instrument + refseq_database_version + uniport_database_version + hgnc_version + raw_data_processing + raw_data_conversion + sequence_database_search + search_database_parameters + phosphosite_localization + ms1_data_analysis + psm_report_generation + cptac_dcc_mzidentml + mzidentml_refseq + mzidentml_uniprot + gene_to_prot + cptac_galaxy_workflows + cptac_galaxy_tools + cdap_reports + cptac_dcc_tools + } + } +query ProtocolQuery($study_id: String!, $source: String!){ + uiProtocol (study_id: $study_id, source: $source ){ + protocol_id + protocol_submitter_id + program_id + program_submitter_id + protocol_name + protocol_date + document_name + quantitation_strategy + experiment_type + label_free_quantitation + labeled_quantitation + isobaric_labeling_reagent + reporter_ion_ms_level + starting_amount + starting_amount_uom + digestion_reagent + alkylation_reagent + enrichment_strategy + enrichment + chromatography_dimensions_count + one_d_chromatography_type + two_d_chromatography_type + fractions_analyzed_count + column_type + amount_on_column + amount_on_column_uom + column_length + column_length_uom + column_inner_diameter + column_inner_diameter_uom + particle_size + particle_size_uom + particle_type + gradient_length + gradient_length_uom + instrument_make + instrument_model + dissociation_type + ms1_resolution + ms2_resolution + dda_topn + normalized_collision_energy + acquistion_type + dia_multiplexing + dia_ims + auxiliary_data + cud_label + analytical_technique + chromatography_instrument_make + chromatography_instrument_model + polarity + reconstitution_solvent + reconstitution_volume + reconstitution_volume_uom + internal_standards + extraction_method + ionization_mode + } + } +query PublicationsQuery($study_id: String!, $source: String!){ + uiPublication (study_id: $study_id, source: $source){ + publication_id + pubmed_id + title + } + } +query FilesCountsQuery($study_id: String!, $source: String!){ + uiFilesCountPerStudy (study_id: $study_id, source: $source ){ + study_submitter_id + file_type + files_count + data_category + } + } +query FilteredClinicalDataPaginated($offset_value: Int, $limit_value: Int, $sort_value: String, $program_name_filter: String!, $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, $analytical_frac_filter: String!, $exp_type_filter: String!, $ethnicity_filter: String!, $race_filter: String!, $gender_filter: String!, $tumor_grade_filter: String!, $sample_type_filter: String!, $acquisition_type_filter: String!, $data_category_filter: String!, $file_type_filter: String!, $access_filter: String!, $downloadable_filter: String!, $case_status_filter: String!, $biospecimen_status_filter: String!, $source: String!){ + getPaginatedUIClinical(offset: $offset_value, limit: $limit_value, sort: $sort_value, program_name: $program_name_filter , project_name: $project_name_filter, + study_name: $study_name_filter, disease_type: $disease_filter, primary_site: $filterValue, analytical_fraction: $analytical_frac_filter, + experiment_type: $exp_type_filter, ethnicity: $ethnicity_filter, race: $race_filter, gender: $gender_filter, + tumor_grade: $tumor_grade_filter, sample_type: $sample_type_filter, acquisition_type: $acquisition_type_filter, data_category: $data_category_filter, file_type: $file_type_filter, access: $access_filter, downloadable: $downloadable_filter, case_status: $case_status_filter, biospecimen_status: $biospecimen_status_filter, source: $source) { + total + uiClinical{ + case_submitter_id + external_case_id + ethnicity + gender + race + morphology + primary_diagnosis + site_of_resection_or_biopsy + tissue_or_organ_of_origin + tumor_grade + tumor_stage + age_at_diagnosis + classification_of_tumor + days_to_recurrence + case_id + disease_type + primary_site + program_name + project_name + status + externalReferences { + reference_resource_shortname + reference_entity_location + } + samples { + sample_id + sample_submitter_id + } + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query FilteredCasesDataPaginated($offset_value: Int, $limit_value: Int, $sort_value: String, $program_name_filter: String!, $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, $analytical_frac_filter: String!, $exp_type_filter: String!, $ethnicity_filter: String!, $race_filter: String!, $gender_filter: String!, $tumor_grade_filter: String!, $sample_type_filter: String!, $acquisition_type_filter: String!, $data_category_filter: String!, $file_type_filter: String!, $access_filter: String!, $downloadable_filter: String!, $biospecimen_status_filter: String!, $case_status_filter: String!, $source: String!){ + getPaginatedUICase(offset: $offset_value, limit: $limit_value, sort: $sort_value, program_name: $program_name_filter , + project_name: $project_name_filter, study_name: $study_name_filter, disease_type: $disease_filter, + primary_site: $filterValue, analytical_fraction: $analytical_frac_filter, experiment_type: $exp_type_filter, + ethnicity: $ethnicity_filter, race: $race_filter, gender: $gender_filter, tumor_grade: $tumor_grade_filter, + sample_type: $sample_type_filter, acquisition_type: $acquisition_type_filter, data_category: $data_category_filter, file_type: $file_type_filter, access: $access_filter, downloadable: $downloadable_filter, biospecimen_status: $biospecimen_status_filter, case_status: $case_status_filter, source: $source) { + total + uiCases{ + aliquot_id + sample_id + case_id + aliquot_submitter_id + aliquot_is_ref + aliquot_status + aliquot_quantity + aliquot_volume + amount + analyte_type + concentration + case_status + sample_status + sample_submitter_id + sample_is_ref + biospecimen_anatomic_site + biospecimen_laterality + composition + current_weight + days_to_collection + days_to_sample_procurement + diagnosis_pathologically_confirmed + freezing_method + initial_weight + intermediate_dimension + longest_dimension + method_of_sample_procurement + pathology_report_uuid + preservation_method + sample_type_id + shortest_dimension + time_between_clamping_and_freezing + time_between_excision_and_freezing + tissue_type + tumor_code + tumor_code_id + tumor_descriptor + case_submitter_id + program_name + project_name + sample_type + disease_type + primary_site + tissue_collection_type + sample_ordinal + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query StudyExperimentalDesign($study_id_value: String, $source: String!) { + uiStudyExperimentalDesign(study_id: $study_id_value, source: $source) { + study_id + study_submitter_id + study_run_metadata_id + study_run_metadata_submitter_id + experiment_number + experiment_type + plex_dataset_name + acquisition_type + number_of_fractions + analyte + polarity + label_free { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_113 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_114 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_115 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_116 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_117 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_118 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_119 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + itraq_121 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_126{ + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_127n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_127c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_128n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_128c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_129n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_129c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_130c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_130n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_131 { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_131c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_132n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_132c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_133n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_133c { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + }, + tmt_134n { + aliquot_id, + aliquot_run_metadata_id, + aliquot_submitter_id + } + } + } +query BiospecimenPerStudy($study_id_value: String, $source: String!) { + uiBiospecimenPerStudy(study_id: $study_id_value, source: $source) { + aliquot_id + sample_id + case_id + aliquot_submitter_id + sample_submitter_id + case_submitter_id + aliquot_is_ref + aliquot_status + case_status + sample_status + project_name + sample_type + disease_type + primary_site + pool + taxon + } + } +query EntityReferenceQueryData($entity_type_filter: String!, $entity_id_filter: String!, $reference_type_filter: String!, $source: String!){ + uiPdcEntityReference(entity_type: $entity_type_filter , entity_id: $entity_id_filter, reference_type: $reference_type_filter, source: $source) { + reference_id + entity_type + entity_id + reference_type + reference_entity_type + reference_entity_alias + reference_resource_name + reference_resource_shortname + reference_entity_location + submitter_id_name + } + } +query FilteredClinicalDataPaginated($offset_value: Int, $limit_value: Int, $sort_value: String, $program_name_filter: String!, $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, $analytical_frac_filter: String!, $exp_type_filter: String!, $ethnicity_filter: String!, $race_filter: String!, $gender_filter: String!, $tumor_grade_filter: String!, $sample_type_filter: String!, $acquisition_type_filter: String!, $data_category_filter: String!, $file_type_filter: String!, $access_filter: String!, $downloadable_filter: String!, $case_status_filter: String!, $biospecimen_status_filter: String!, $source: String!){ + getPaginatedUIClinical(offset: $offset_value, limit: $limit_value, sort: $sort_value, program_name: $program_name_filter , project_name: $project_name_filter, + study_name: $study_name_filter, disease_type: $disease_filter, primary_site: $filterValue, analytical_fraction: $analytical_frac_filter, + experiment_type: $exp_type_filter, ethnicity: $ethnicity_filter, race: $race_filter, gender: $gender_filter, + tumor_grade: $tumor_grade_filter, sample_type: $sample_type_filter, acquisition_type: $acquisition_type_filter, data_category: $data_category_filter, file_type: $file_type_filter, access: $access_filter, downloadable: $downloadable_filter, case_status: $case_status_filter, biospecimen_status: $biospecimen_status_filter, source: $source) { + total + uiClinical{ + case_submitter_id + external_case_id + ethnicity + gender + race + morphology + primary_diagnosis + site_of_resection_or_biopsy + tissue_or_organ_of_origin + tumor_grade + tumor_stage + age_at_diagnosis + classification_of_tumor + days_to_recurrence + case_id + disease_type + primary_site + program_name + project_name + status + cause_of_death + days_to_birth + days_to_death + vital_status + year_of_birth + year_of_death + age_at_index + premature_at_birth + weeks_gestation_at_birth + age_is_obfuscated + cause_of_death_source + occupation_duration_years + country_of_residence_at_enrollment + days_to_last_follow_up + days_to_last_known_disease_status + last_known_disease_status + progression_or_recurrence + prior_malignancy + ajcc_clinical_m + ajcc_clinical_n + ajcc_clinical_stage + ajcc_clinical_t + ajcc_pathologic_m + ajcc_pathologic_n + ajcc_pathologic_stage + ajcc_pathologic_t + ajcc_staging_system_edition + ann_arbor_b_symptoms + ann_arbor_clinical_stage + ann_arbor_extranodal_involvement + ann_arbor_pathologic_stage + best_overall_response + burkitt_lymphoma_clinical_variant + circumferential_resection_margin + colon_polyps_history + days_to_best_overall_response + days_to_diagnosis + days_to_hiv_diagnosis + days_to_new_event + figo_stage + hiv_positive + hpv_positive_type + hpv_status + iss_stage + laterality + ldh_level_at_diagnosis + ldh_normal_range_upper + lymph_nodes_positive + lymphatic_invasion_present + method_of_diagnosis + peripancreatic_lymph_nodes_positive + peripancreatic_lymph_nodes_tested + supratentorial_localization + tumor_confined_to_organ_of_origin + tumor_focality + tumor_regression_grade + vascular_invasion_type + wilms_tumor_histologic_subtype + breslow_thickness + gleason_grade_group + igcccg_stage + international_prognostic_index + largest_extrapelvic_peritoneal_focus + masaoka_stage + new_event_anatomic_site + new_event_type + overall_survival + perineural_invasion_present + prior_treatment + progression_free_survival + progression_free_survival_event + residual_disease + vascular_invasion_present + year_of_diagnosis + icd_10_code + synchronous_malignancy + metastasis_at_diagnosis + metastasis_at_diagnosis_site + mitosis_karyorrhexis_index + non_nodal_regional_disease + non_nodal_tumor_deposits + ovarian_specimen_status + ovarian_surface_involvement + percent_tumor_invasion + peritoneal_fluid_cytological_status + primary_gleason_grade + secondary_gleason_grade + weiss_assessment_score + adrenal_hormone + ann_arbor_b_symptoms_described + diagnosis_is_primary_disease + eln_risk_classification + figo_staging_edition_year + gleason_grade_tertiary + gleason_patterns_percent + margin_distance + margins_involved_site + pregnant_at_diagnosis + satellite_nodule_present + sites_of_involvement + tumor_depth + who_cns_grade + who_nte_grade + diagnosis_uuid + anaplasia_present + anaplasia_present_type + child_pugh_classification + cog_liver_stage + cog_neuroblastoma_risk_group + cog_renal_stage + cog_rhabdomyosarcoma_risk_group + enneking_msts_grade + enneking_msts_metastasis + enneking_msts_stage + enneking_msts_tumor_site + esophageal_columnar_dysplasia_degree + esophageal_columnar_metaplasia_present + first_symptom_prior_to_diagnosis + gastric_esophageal_junction_involvement + goblet_cells_columnar_mucosa_present + gross_tumor_weight + inpc_grade + inpc_histologic_group + inrg_stage + inss_stage + irs_group + irs_stage + ishak_fibrosis_score + lymph_nodes_tested + medulloblastoma_molecular_classification + externalReferences { + reference_resource_shortname + reference_entity_location + } + exposures { + exposure_id + exposure_submitter_id + alcohol_days_per_week + alcohol_drinks_per_day + alcohol_history + alcohol_intensity + asbestos_exposure + cigarettes_per_day + coal_dust_exposure + environmental_tobacco_smoke_exposure + pack_years_smoked + radon_exposure + respirable_crystalline_silica_exposure + smoking_frequency + time_between_waking_and_first_smoke + tobacco_smoking_onset_year + tobacco_smoking_quit_year + tobacco_smoking_status + type_of_smoke_exposure + type_of_tobacco_used + years_smoked + age_at_onset + alcohol_type + exposure_duration + exposure_duration_years + exposure_type + marijuana_use_per_week + parent_with_radiation_exposure + secondhand_smoke_as_child + smokeless_tobacco_quit_age + tobacco_use_per_day + } + follow_ups { + follow_up_id + follow_up_submitter_id + adverse_event + adverse_event_grade + aids_risk_factors + barretts_esophagus_goblet_cells_present + bmi + body_surface_area + cause_of_response + cd4_count + cdc_hiv_risk_factors + comorbidity + comorbidity_method_of_diagnosis + days_to_adverse_event + days_to_comorbidity + days_to_follow_up + days_to_imaging + days_to_progression + days_to_progression_free + days_to_recurrence + diabetes_treatment_type + disease_response + dlco_ref_predictive_percent + ecog_performance_status + evidence_of_recurrence_type + eye_color + fev1_ref_post_bronch_percent + fev1_ref_pre_bronch_percent + fev1_fvc_pre_bronch_percent + fev1_fvc_post_bronch_percent + haart_treatment_indicator + height + hepatitis_sustained_virological_response + history_of_tumor + history_of_tumor_type + hiv_viral_load + hormonal_contraceptive_type + hormonal_contraceptive_use + hormone_replacement_therapy_type + hpv_positive_type + hysterectomy_margins_involved + hysterectomy_type + imaging_result + imaging_type + immunosuppressive_treatment_type + karnofsky_performance_status + menopause_status + nadir_cd4_count + pancreatitis_onset_year + pregnancy_outcome + procedures_performed + progression_or_recurrence + progression_or_recurrence_anatomic_site + progression_or_recurrence_type + recist_targeted_regions_number + recist_targeted_regions_sum + reflux_treatment_type + risk_factor + risk_factor_treatment + scan_tracer_used + undescended_testis_corrected + undescended_testis_corrected_age + undescended_testis_corrected_laterality + undescended_testis_corrected_method + undescended_testis_history + undescended_testis_history_laterality + viral_hepatitis_serologies + weight + } + treatments { + treatment_id + treatment_submitter_id + days_to_treatment_end + days_to_treatment_start + initial_disease_status + regimen_or_line_of_therapy + therapeutic_agents + treatment_anatomic_site + treatment_effect + treatment_intent_type + treatment_or_therapy + treatment_outcome + treatment_type + chemo_concurrent_to_radiation + number_of_cycles + reason_treatment_ended + route_of_administration + treatment_arm + treatment_dose + treatment_dose_units + treatment_effect_indicator + treatment_frequency + } + samples { + sample_id + sample_submitter_id + annotation + } + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query CasesData{ + uiStudy { + submitter_id_name + program_name + project_name + disease_type + primary_site + analytical_fraction + experiment_type + num_raw + num_mzml + num_prot + num_prot_assem + num_psm + } + } +query ProteinQuery($protein_name: String!){ + uiProtein(protein: $protein_name){ + gene_name + NCBI_gene_id + alias + authority + description + organism + chromosome + locus + proteins + assays + spectral_counts { + project_submitter_id + plex + spectral_count + distinct_peptide + unshared_peptide + } + } + } +query ProteinQuery($gene_name: String!, $source: String!){ + uiGeneSpectralCount(gene_name: $gene_name, source: $source){ + gene_id + gene_name + ncbi_gene_id + alias + authority + description + organism + chromosome + locus + proteins + assays + spectral_counts { + project_submitter_id + study_submitter_id + plex + spectral_count + distinct_peptide + unshared_peptide + } + } + } +query ProteinQuery($gene_name: String, $gene_id: String $source: String!){ + uiGeneSpectralCount(gene_name: $gene_name, gene_id: $gene_id, source: $source){ + gene_id + gene_name + ncbi_gene_id + alias + authority + description + organism + chromosome + locus + proteins + assays + spectral_counts { + project_submitter_id + study_submitter_id + plex + spectral_count + distinct_peptide + unshared_peptide + } + } + } +query aliquotSpectralCountQuery($gene_name:String!, $offset_param: Int, $limit_param: Int, $source: String!){ + getPaginatedUIGeneAliquotSpectralCount(gene_name: $gene_name, offset: $offset_param, limit: $limit_param, source: $source){ + total + uiGeneAliquotSpectralCounts { + aliquot_id + plex + label + submitter_id_name + experiment_type + spectral_count + distinct_peptide + unshared_peptide + precursor_area + log2_ratio + unshared_precursor_area + unshared_log2_ratio + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query studySpectralCountQuery($gene_name:String!, $offset_param: Int, $limit_param: Int, $source: String!){ + getPaginatedUIGeneStudySpectralCount(gene_name: $gene_name, offset: $offset_param, limit: $limit_param, source: $source){ + total + uiGeneStudySpectralCounts { + submitter_id_name + experiment_type + spectral_count + distinct_peptide + unshared_peptide + aliquots_count + plexes_count + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query PTMDataByGeneQuery($gene_name:String!, $offset_param: Int, $limit_param: Int, $source: String!){ + getPaginatedUIPtm(gene_name: $gene_name offset: $offset_param limit: $limit_param, source: $source){ + total + uiPtm { + ptm_type + site + peptide + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query PTMDataByGeneQuery($gene_name: String, $gene_id:String, $offset_param: Int, $limit_param: Int, $source: String!){ + getPaginatedUIPtm(gene_name: $gene_name gene_id: $gene_id offset: $offset_param limit: $limit_param, source: $source){ + total + uiPtm { + ptm_type + site + peptide + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query GeneStudyCountQuery($gene_id: String!){ + geneStudyCount(gene_id: $gene_id) + } +query allTissueSites{ + uiTissueSiteCaseCount { + tissue_or_organ_of_origin + cases_count + } + } +query allStats{ + uiPdcDataStats{ + program, + study, + spectra, + data_label, + protein, + project, + peptide, + data_size, + data_label, + data_file + } + } +query allDiseases{ + uiDiseasesAvailable { + disease_type + tissue_or_organ_of_origin + project_submitter_id + cases_count + } + } +query Programs{ + uiAllPrograms { + program_submitter_id + name + sponsor + start_date + end_date + program_manager + projects { + project_submitter_id + } + } + } +query sunburstChartData{ + uiSunburstChart { + project_submitter_id + tissue_or_organ_of_origin + disease_type + sample_type + cases_count + } + } +query HumanBodyPrimarySiteData { + uiPrimarySiteCaseCount{ + major_primary_site + cases_count + primarySites + } + } +query CaseSearchQuery($case_name: String!){ + caseSearch(name: $case_name){ + searchCases { + record_type + name + case_id + } + } + } +query GeneSearchQuery($gene_name: String!){ + geneSearch(name: $gene_name){ + genes { + record_type + name + description + gene_id + ncbi_gene_id + alias + } + } + } +query ProteinSearchQuery($protein_name: String!){ + proteinSearch(name: $protein_name){ + genesWithProtein { + record_type + name + description + gene_id + ncbi_gene_id + proteins + } + } + } +query StudySearchQuery($study_name: String!){ + studySearch(name: $study_name){ + studies{ + record_type + name + submitter_id_name + study_id + study_submitter_id + pdc_study_id + } + } + } +query StudySearchByPDCStudyIdQuery($pdc_study_id: String!){ + studySearchByPDCStudyId(pdc_study_id: $pdc_study_id){ + studies { + record_type + name + submitter_id_name + study_id + study_submitter_id + pdc_study_id + } + } + } +query StudySearchByExternalRefQuery($reference_entity_alias: String!){ + studySearchByExternalId(reference_entity_alias: $reference_entity_alias){ + studies { + record_type + name + submitter_id_name + study_id + study_submitter_id + pdc_study_id + } + } + } +query AliquotSearchQuery($aliquot_submitter_id: String!){ + aliquotSearch(name: $aliquot_submitter_id){ + searchAliquots{ + aliquot_id + aliquot_submitter_id + } + } + } +query studyByUUIDQuery($study_id: String!, $study_submitter_id: String!, $source: String!){ + uiStudySummary(study_id: $study_id, study_submitter_id: $study_submitter_id, source: $source){ + study_submitter_id + study_shortname + } + } +query studyByUUIDQuery($study_id: String!, $pdc_study_id: String!, $source: String!){ + uiStudySummary(study_id: $study_id, pdc_study_id: $pdc_study_id, source: $source){ + study_submitter_id + study_name + pdc_study_id + study_id + } + } +query caseByUUIDQuery($case_id: String!, $source: String!){ + uiCaseSummary(case_id: $case_id, source: $source){ + case_submitter_id + case_id + } + } +query CaseSummaryDataQuery($case_id: String!, $case_submitter_id: String!, $source: String!){ + uiCase (case_id: $case_id, case_submitter_id: $case_submitter_id, source: $source) { + case_id + case_submitter_id + aliquot_id + sample_id + project_name + program_name + sample_type + disease_type + primary_site + aliquot_submitter_id + sample_submitter_id + } + } +query SampleDataQuery($sample_id: String!, $source: String!){ + uiSampleSummary (sample_id: $sample_id, source: $source) { + case_submitter_id + sample_submitter_id + sample_id + } + } +query SampleDataQuery($sample_submitter_id: String!, $source: String!){ + uiSampleSummary (sample_submitter_id: $sample_submitter_id, source: $source) { + case_submitter_id + sample_submitter_id + sample_id + } + } +query AliquotDataQuery($aliquot_id: String!, $source: String!){ + uiAliquotSummary (aliquot_id: $aliquot_id, source: $source) { + case_submitter_id + aliquot_submitter_id + aliquot_id + } + } +query AliquotDataQuery($aliquot_submitter_id: String!, $source: String!){ + uiAliquotSummary (aliquot_submitter_id: $aliquot_submitter_id, source: $source) { + case_submitter_id + aliquot_submitter_id + aliquot_id + } + } +query ObjectSearchedQuery($type: String!, $parameterType: String!, $parameterValue: String!){ + objectSearched (type: $type, parameterType: $parameterType, parameterValue: $parameterValue) + } +query uiDataVersionSoftwareVersion { + uiDataVersionSoftwareVersion { + data_release + build_tag + } + } +query ProteinQuery($protein_name: String!){ + uiProtein(protein: $protein_name){ + gene_name + NCBI_gene_id + authority + description + organism + chromosome + locus + proteins + assays + spectral_counts { + project_submitter_id + plex + spectral_count + distinct_peptide + unshared_peptide + } + } + } +query ProteinQuery($gene_name: String, $gene_id: String){ + uiGeneSpectralCount(gene_name: $gene_name, gene_id: $gene_id){ + gene_id + gene_name + ncbi_gene_id + alias + authority + description + organism + chromosome + locus + proteins + assays + spectral_counts { + project_submitter_id + study_submitter_id + plex + spectral_count + distinct_peptide + unshared_peptide + } + } + } +query aliquotSpectralCountQuery($gene_name:String, $gene_id:String, $offset_param: Int, $limit_param: Int, $program_name_filter: String!, + $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, + $analytical_frac_filter: String!, $exp_type_filter: String!, $ethnicity_filter: String!, $race_filter: String!, + $gender_filter: String!, $tumor_grade_filter: String!, $sample_type_filter: String!, $acquisition_type_filter: String!){ + getPaginatedUIGeneAliquotSpectralCountFiltered(gene_name: $gene_name, gene_id: $gene_id, offset: $offset_param, limit: $limit_param, program_name: $program_name_filter , + project_name: $project_name_filter, study_name: $study_name_filter, disease_type: $disease_filter, primary_site: $filterValue, + analytical_fraction: $analytical_frac_filter, experiment_type: $exp_type_filter, ethnicity: $ethnicity_filter, race: $race_filter, + gender: $gender_filter, tumor_grade: $tumor_grade_filter, sample_type: $sample_type_filter, acquisition_type: $acquisition_type_filter){ + total + uiGeneAliquotSpectralCounts { + aliquot_id + plex + label + submitter_id_name + pdc_study_id + experiment_type + spectral_count + distinct_peptide + unshared_peptide + precursor_area + log2_ratio + unshared_precursor_area + unshared_log2_ratio + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query studySpectralCountQuery($gene_name:String, $gene_id:String, $offset_param: Int, $limit_param: Int, $program_name_filter: String!, + $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, + $analytical_frac_filter: String!, $exp_type_filter: String!, $ethnicity_filter: String!, $race_filter: String!, + $gender_filter: String!, $tumor_grade_filter: String!, $sample_type_filter: String!, $acquisition_type_filter: String!){ + getPaginatedUIGeneStudySpectralCountFiltered(gene_name: $gene_name, gene_id: $gene_id, offset: $offset_param, limit: $limit_param, program_name: $program_name_filter , + project_name: $project_name_filter, study_name: $study_name_filter, disease_type: $disease_filter, primary_site: $filterValue, + analytical_fraction: $analytical_frac_filter, experiment_type: $exp_type_filter, ethnicity: $ethnicity_filter, race: $race_filter, + gender: $gender_filter, tumor_grade: $tumor_grade_filter, sample_type: $sample_type_filter, acquisition_type: $acquisition_type_filter){ + total + uiGeneStudySpectralCounts { + pdc_study_id + submitter_id_name + experiment_type + spectral_count + distinct_peptide + unshared_peptide + aliquots_count + plexes_count + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query PTMDataByGeneQuery($gene_name:String!, $gene_id:String, $offset_param: Int, $limit_param: Int){ + getPaginatedUIPtm(gene_name: $gene_name gene_id: $gene_id offset: $offset_param limit: $limit_param){ + total + uiPtm { + ptm_type + site + peptide + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query GeneStudyData($gene_name: String!){ + uiGeneStudySpectralCount(gene_name: $gene_name) { + study_submitter_id + submitter_id_name + } + } + +query GeneSearchQuery($gene_name: String!){ + geneSearch(name: $gene_name){ + genes { + record_type + name + description + alias + } + } + } +query FiltersData($program_name_filter: String!, $project_name_filter: String!, $study_name_filter: String!, $disease_filter: String!, $filterValue: String!, $analytical_frac_filter: String!, $exp_type_filter: String!, $ethnicity_filter: String!, $race_filter: String!, $gender_filter: String!, $tumor_grade_filter: String!, $sample_type_filter: String!, $acquisition_type_filter: String!, $data_category_filter: String!, $file_type_filter: String!, $access_filter: String!, $downloadable_filter: String!, $biospecimen_status_filter: String!, $case_status_filter: String!){ + uiFilters(program_name: $program_name_filter , project_name: $project_name_filter, + study_name: $study_name_filter, disease_type: $disease_filter, primary_site: $filterValue, analytical_fraction: $analytical_frac_filter, + experiment_type: $exp_type_filter, ethnicity: $ethnicity_filter, race: $race_filter, gender: $gender_filter, tumor_grade: $tumor_grade_filter, + sample_type: $sample_type_filter, acquisition_type: $acquisition_type_filter, data_category: $data_category_filter, file_type: $file_type_filter, access: $access_filter, downloadable: $downloadable_filter, biospecimen_status: $biospecimen_status_filter, case_status: $case_status_filter) { + project_name { + filterName + filterValue + } + primary_site { + filterName + filterValue + } + program_name { + filterName + filterValue + } + disease_type { + filterName + filterValue + } + analytical_fraction { + filterName + filterValue + } + experiment_type { + filterName + filterValue + } + acquisition_type { + filterName + filterValue + } + submitter_id_name { + filterName + filterValue + } + sample_type { + filterName + filterValue + } + ethnicity { + filterName + filterValue + } + race { + filterName + filterValue + } + gender { + filterName + filterValue + } + tumor_grade { + filterName + filterValue + } + data_category { + filterName + filterValue + } + file_type { + filterName + filterValue + } + access { + filterName + filterValue + } + downloadable { + filterName + filterValue + } + biospecimen_status { + filterName + filterValue + } + case_status { + filterName + filterValue + } + } + } + +query CasesData{ + uiStudy { + submitter_id_name + program_name + project_name + disease_type + primary_site + analytical_fraction + experiment_type + num_raw + num_mzml + num_prot + num_prot_assem + num_psm + } + } +query FiltersData { + uiFilters{ + project_name { + filterName + filterValue + } + primary_site { + filterName + filterValue + } + program_name { + filterName + filterValue + } + disease_type { + filterName + filterValue + } + analytical_fraction { + filterName + filterValue + } + experiment_type { + filterName + filterValue + } + acquisition_type { + filterName + filterValue + } + submitter_id_name { + filterName + filterValue + } + sample_type { + filterName + filterValue + } + ethnicity { + filterName + filterValue + } + race { + filterName + filterValue + } + gender { + filterName + filterValue + } + tumor_grade { + filterName + filterValue + } + data_category { + filterName + filterValue + } + file_type { + filterName + filterValue + } + access { + filterName + filterValue + } + } + } +query FilterdPaginatedPublicationsQuery ($offset_value: Int, $limit_value: Int, $year_filter: String!, $disease_type_filter: String!, $program_filter: String!, $pubmedid_filter: String!) { + getPaginatedUIPublication(offset: $offset_value, limit: $limit_value, year: $year_filter, disease_type: $disease_type_filter, program: $program_filter, pubmed_id: $pubmedid_filter) { + total + uiPublication { + publication_id + program_name + pubmed_id + doi + author + title + journal + journal_url + year + abstract + citation + studies { + study_id + pdc_study_id + submitter_id_name + } + disease_types + supplementary_data + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query PubmedIDSearchQuery($search_term: String!){ + getPaginatedUIPublication(pubmed_id: $search_term) { + uiPublication { + pubmed_id + title + } + } + } +query allPublicationsFilters{ + getUIPublicationFilters { + disease_types + years + programs + } + } +query UILegacyStudyQuery($study_name: String!){ + uiLegacyStudies(pdc_study_id: $study_name) { + study_id + submitter_id_name + study_submitter_id + pdc_study_id + study_description + project_submitter_id + analytical_fraction + experiment_type + embargo_date + sort_order + supplementaryFilesCount { + data_category + file_type + files_count + } + nonSupplementaryFilesCount { + data_category + file_type + files_count + } + publications { + publication_id + pubmed_id + doi + author + title + journal + journal_url + year + abstract + citation + } + } + } +query PublicationsQuery($study_id: String!){ + uiPublication (study_id: $study_id ){ + publication_id + pubmed_id + title + } + } +query FilesCountsQuery($study_id: String!){ + uiFilesCountPerStudy (study_id: $study_id ){ + study_submitter_id + file_type + files_count + data_category + } + } +query LegacyDataEntityReferenceQueryData($entity_id_filter: String!, $reference_type_filter: String!){ + uiLegacyStudyReference(entity_id: $entity_id_filter, reference_type: $reference_type_filter) { + reference_id + entity_type + entity_id + reference_type + reference_entity_type + reference_entity_alias + reference_resource_name + reference_resource_shortname + reference_entity_location + submitter_id_name + } + } +query FilterdPaginatedPublicationsQuery ($offset_value: Int, $limit_value: Int, $year_filter: String!, $disease_type_filter: String!, $program_filter: String!) { + getPaginatedUIPublication(offset: $offset_value, limit: $limit_value, year: $year_filter, disease_type: $disease_type_filter, program: $program_filter) { + total + uiPublication { + publication_id + pubmed_id + doi + author + title + journal + journal_url + year + abstract + citation + studies { + pdc_study_id + submitter_id_name + } + disease_types + supplementary_data + } + pagination { + count + sort + from + page + total + pages + size + } + } + } +query allLegacyStudies{ + uiLegacyStudies { + study_id + submitter_id_name + study_submitter_id + pdc_study_id + study_description + project_submitter_id + analytical_fraction + experiment_type + sort_order + embargo_date + supplementaryFilesCount { + data_category + file_type + files_count + } + nonSupplementaryFilesCount{ + data_category + file_type + files_count + } + publications { + publication_id + pubmed_id + doi + author + title + journal + journal_url + year + abstract + citation + } + } + } +query FilterdHeatmapsQuery ($sort_value: String!, $primary_sites_filter: String!, $disease_type_filter: String!, $analytical_fractions_filter: String!) { + uiHeatmapStudies ( sort: $sort_value, primary_site: $primary_sites_filter, disease_type: $disease_type_filter, analytical_fraction: $analytical_fractions_filter) { + study_id + study_submitter_id + submitter_id_name + pdc_study_id + study_description + program_name + project_name + analytical_fraction + primary_site + disease_type + experiment_type + embargo_date + heatmapFiles + } + } +query allHeatmapsFilters{ + getUIHeatmapFilters { + filterCount + disease_types + primary_sites + analytical_fractions + } + } +query StudySearchByExternalRefQuery($reference_entity_alias: String!){ + studySearchByExternalId(reference_entity_alias: $reference_entity_alias){ + studies { + record_type + name + submitter_id_name + study_id + study_submitter_id + pdc_study_id + } + } + } +query PubmedIDSearchQuery { + getUIPancancerPublications { + publication_id + pubmed_id + doi + author + title + journal + journal_url + year + abstract + citation + files { + file_id + file_name + data_category + description + downloadable + characterization + cohorts + related_publications + related_studies + } + studies { + study_id + submitter_id_name + pdc_study_id + } + } + } +query PancancerFilesQuery { + getAllUIPancancerFiles { + file_id + file_name + data_category + description + downloadable + characterization + cohorts + related_publications + related_studies + } + } diff --git a/chat-ui/beaker-biome/src/beaker_biome/integrations/pdc_regex.py b/chat-ui/beaker-biome/src/beaker_biome/integrations/pdc_regex.py new file mode 100644 index 0000000..5a74a5a --- /dev/null +++ b/chat-ui/beaker-biome/src/beaker_biome/integrations/pdc_regex.py @@ -0,0 +1,30 @@ +#!/usr/bin/env python + +# usage: get the PDC js webpack bundle from the website's network request, +# run it through prettifier/any other JS formatter to ensure the string templates are easy +# to extract with regex, then run +# $ python pdc_regex.py > pdc_schema.graphql +# to get a list of queries isolated. + +import sys +import re + +def main(): + if len(sys.argv) < 2: + print("usage: pdc_regex.py [filename]") + return + try: + with open(sys.argv[1], 'r') as file: + content = file.read() + except FileNotFoundError: + print(f"'{sys.argv[1]}': file not found") + return + except Exception as e: + print(f"{e}") + return + pattern = re.compile(r'`\n\s*(query [^`]+)', re.MULTILINE | re.IGNORECASE) + results = [match.group(1) for match in re.finditer(pattern, content)] + [print(result) for result in results] + +if __name__ == "__main__": + main() From f51f297dc087a104566cdf53ecab33a40ffcfc60 Mon Sep 17 00:00:00 2001 From: Satchel Baldwin Date: Wed, 30 Oct 2024 16:47:47 -0500 Subject: [PATCH 2/2] add pdc support --- .../src/beaker_biome/biome/api_agent.yaml | 23 ++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-) diff --git a/chat-ui/beaker-biome/src/beaker_biome/biome/api_agent.yaml b/chat-ui/beaker-biome/src/beaker_biome/biome/api_agent.yaml index 9db277f..3c6775d 100644 --- a/chat-ui/beaker-biome/src/beaker_biome/biome/api_agent.yaml +++ b/chat-ui/beaker-biome/src/beaker_biome/biome/api_agent.yaml @@ -94,7 +94,6 @@ apis: documentation_file: "gdc.md" cache: key: "api_assistant_gdc_faceted" - # for brevity, we define `disease_types` below and use it to avoid repetition. gpt_additional_pass: | You will be given python code to query the GDC API. @@ -1890,3 +1889,25 @@ apis: - complete genomics - illumina methylation epic v2 - genechip u133 plus 2.0 + pdc: + name: "Proteomics Data Commons" + description: | + The objectives of the National Cancer Institute’s Proteomic Data Commons (PDC) are (1) to make cancer-related proteomic datasets easily accessible to the public, and (2) facilitate direct multiomics integration in support of precision medicine through interoperability with accompanying data resources (genomic and medical image datasets). + The PDC was developed to advance our understanding of how proteins help to shape the risk, diagnosis, development, progression, and treatment of cancer. In-depth analysis of proteomic data allows us to study both how and why cancer develops and to devise ways of personalizing treatment for patients using precision medicine. + The PDC is one of several repositories within the NCI Cancer Research Data Commons (CRDC), a secure cloud-based infrastructure featuring diverse data sets and innovative analytic tools – all designed to advance data-driven scientific discovery. The CRDC enables researchers to link proteomic data with other data sets (e.g., genomic and imaging data) and to submit, collect, analyze, store, and share data throughout the cancer data ecosystem. + Access to highly curated and standardized biospecimen, clinical and proteomic data with direct integration of accompanying data resources (genomic and medical image datasets). + + Uses: + * Intuitive interface to filter, query, search, visualize and download the data and metadata. + * A common data harmonization pipeline to uniformly analyze all PDC data and provide advanced visualization of the quantitative information. + * Cloud based (Amazon Web Services) infrastructure facilitates interoperability with AWS based data analysis tools and platforms natively. + * Application programming interface (API) provides cloud-agnostic data access and allows third parties to extend the functionality beyond the PDC. + * A highly structured workspace that serves as a private user data store and also data submission portal. + * Distributes controlled access data, such as the patient-specific protein fasta sequence databases, with dbGaP authorization and eRA Commons authentication. + documentation_file: "pdc_schema.graphql" + cache: + key: "api_assistant_pdc_graphql" + additional_cache_body: | + This API is through GraphQL. You will be provided the schema. + All requests go to the following URL: https://pdc.cancer.gov/graphql + Make all GraphQL requests to that URL.