Datasets:
Formats:
parquet
Languages:
English
Size:
10M - 100M
Tags:
biology
chemistry
drug-discovery
clinical-trials
protein-protein-interaction
gene-essentiality
License:
| """Tests for NegBioDB GE (Gene Essentiality) database layer. | |
| Tests migration, connection, table creation, and pair aggregation. | |
| """ | |
| import sqlite3 | |
| import tempfile | |
| from pathlib import Path | |
| import pytest | |
| from negbiodb_depmap.depmap_db import ( | |
| create_ge_database, | |
| get_connection, | |
| refresh_all_ge_pairs, | |
| run_ge_migrations, | |
| ) | |
| MIGRATIONS_DIR = Path(__file__).parent.parent / "migrations_depmap" | |
| def tmp_db(tmp_path): | |
| """Create a temporary GE database with all migrations applied.""" | |
| db_path = tmp_path / "test_ge.db" | |
| run_ge_migrations(db_path, MIGRATIONS_DIR) | |
| return db_path | |
| def conn(tmp_db): | |
| """Get a connection to the temporary GE database.""" | |
| c = get_connection(tmp_db) | |
| yield c | |
| c.close() | |
| # ── Migration tests ─────────────────────────────────────────────────── | |
| class TestMigrations: | |
| def test_migration_creates_all_tables(self, conn): | |
| """All 10 expected tables should exist after migration.""" | |
| tables = { | |
| row[0] | |
| for row in conn.execute( | |
| "SELECT name FROM sqlite_master WHERE type='table'" | |
| ).fetchall() | |
| } | |
| expected = { | |
| "schema_migrations", | |
| "dataset_versions", | |
| "genes", | |
| "cell_lines", | |
| "ge_screens", | |
| "ge_negative_results", | |
| "gene_cell_pairs", | |
| "ge_split_definitions", | |
| "ge_split_assignments", | |
| "prism_compounds", | |
| "prism_sensitivity", | |
| } | |
| assert expected.issubset(tables), f"Missing tables: {expected - tables}" | |
| def test_migration_version_recorded(self, conn): | |
| """Migration 001 should be recorded in schema_migrations.""" | |
| versions = { | |
| row[0] | |
| for row in conn.execute( | |
| "SELECT version FROM schema_migrations" | |
| ).fetchall() | |
| } | |
| assert "001" in versions | |
| def test_migration_idempotent(self, tmp_db): | |
| """Running migrations twice should not fail or duplicate.""" | |
| applied = run_ge_migrations(tmp_db, MIGRATIONS_DIR) | |
| assert applied == [], "No new migrations expected on second run" | |
| def test_create_ge_database(self, tmp_path): | |
| """create_ge_database convenience wrapper should work.""" | |
| db_path = tmp_path / "convenience.db" | |
| result = create_ge_database(db_path, MIGRATIONS_DIR) | |
| assert result == db_path | |
| assert db_path.exists() | |
| # ── Connection tests ────────────────────────────────────────────────── | |
| class TestConnection: | |
| def test_wal_mode(self, conn): | |
| mode = conn.execute("PRAGMA journal_mode").fetchone()[0] | |
| assert mode == "wal" | |
| def test_foreign_keys_enabled(self, conn): | |
| fk = conn.execute("PRAGMA foreign_keys").fetchone()[0] | |
| assert fk == 1 | |
| # ── Schema constraint tests ────────────────────────────────────────── | |
| class TestSchemaConstraints: | |
| def test_gene_entrez_unique(self, conn): | |
| """Entrez ID should be unique.""" | |
| conn.execute( | |
| "INSERT INTO genes (entrez_id, gene_symbol) VALUES (7157, 'TP53')" | |
| ) | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| "INSERT INTO genes (entrez_id, gene_symbol) VALUES (7157, 'TP53_DUP')" | |
| ) | |
| def test_cell_line_model_id_unique(self, conn): | |
| """Model ID should be unique.""" | |
| conn.execute( | |
| "INSERT INTO cell_lines (model_id, ccle_name) VALUES ('ACH-000001', 'A549_LUNG')" | |
| ) | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| "INSERT INTO cell_lines (model_id, ccle_name) VALUES ('ACH-000001', 'DUP')" | |
| ) | |
| def test_confidence_tier_check(self, conn): | |
| """Invalid confidence tier should fail.""" | |
| conn.execute("INSERT INTO genes (entrez_id, gene_symbol) VALUES (1, 'A')") | |
| conn.execute("INSERT INTO cell_lines (model_id) VALUES ('ACH-000001')") | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, evidence_type, confidence_tier, | |
| source_db, source_record_id, extraction_method) | |
| VALUES (1, 1, 'crispr_nonessential', 'platinum', | |
| 'depmap', 'test', 'score_threshold')""" | |
| ) | |
| def test_evidence_type_check(self, conn): | |
| """Invalid evidence type should fail.""" | |
| conn.execute("INSERT INTO genes (entrez_id, gene_symbol) VALUES (1, 'A')") | |
| conn.execute("INSERT INTO cell_lines (model_id) VALUES ('ACH-000001')") | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, evidence_type, confidence_tier, | |
| source_db, source_record_id, extraction_method) | |
| VALUES (1, 1, 'invalid_type', 'bronze', | |
| 'depmap', 'test', 'score_threshold')""" | |
| ) | |
| def test_screen_type_check(self, conn): | |
| """Screen type must be crispr or rnai.""" | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| """INSERT INTO ge_screens | |
| (source_db, depmap_release, screen_type) | |
| VALUES ('depmap', '25Q3', 'sirna')""" | |
| ) | |
| def test_negative_result_dedup(self, conn): | |
| """Duplicate gene-cell_line-screen-source should be rejected.""" | |
| conn.execute("INSERT INTO genes (entrez_id, gene_symbol) VALUES (1, 'A')") | |
| conn.execute("INSERT INTO cell_lines (model_id) VALUES ('ACH-000001')") | |
| conn.execute( | |
| """INSERT INTO ge_screens (source_db, depmap_release, screen_type) | |
| VALUES ('depmap', '25Q3', 'crispr')""" | |
| ) | |
| conn.commit() | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, screen_id, evidence_type, confidence_tier, | |
| source_db, source_record_id, extraction_method) | |
| VALUES (1, 1, 1, 'crispr_nonessential', 'bronze', | |
| 'depmap', 'r1', 'score_threshold')""" | |
| ) | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, screen_id, evidence_type, confidence_tier, | |
| source_db, source_record_id, extraction_method) | |
| VALUES (1, 1, 1, 'crispr_nonessential', 'bronze', | |
| 'depmap', 'r2', 'score_threshold')""" | |
| ) | |
| def test_prism_broad_id_unique(self, conn): | |
| """PRISM broad_id should be unique.""" | |
| conn.execute( | |
| "INSERT INTO prism_compounds (broad_id, name) VALUES ('BRD-K001', 'Drug1')" | |
| ) | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| "INSERT INTO prism_compounds (broad_id, name) VALUES ('BRD-K001', 'Drug2')" | |
| ) | |
| def test_foreign_key_gene(self, conn): | |
| """FK from ge_negative_results to genes should be enforced.""" | |
| conn.execute("INSERT INTO cell_lines (model_id) VALUES ('ACH-000001')") | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, evidence_type, confidence_tier, | |
| source_db, source_record_id, extraction_method) | |
| VALUES (999, 1, 'crispr_nonessential', 'bronze', | |
| 'depmap', 'test', 'score_threshold')""" | |
| ) | |
| def test_foreign_key_cell_line(self, conn): | |
| """FK from ge_negative_results to cell_lines should be enforced.""" | |
| conn.execute("INSERT INTO genes (entrez_id, gene_symbol) VALUES (1, 'A')") | |
| conn.commit() | |
| with pytest.raises(sqlite3.IntegrityError): | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, evidence_type, confidence_tier, | |
| source_db, source_record_id, extraction_method) | |
| VALUES (1, 999, 'crispr_nonessential', 'bronze', | |
| 'depmap', 'test', 'score_threshold')""" | |
| ) | |
| # ── Pair aggregation tests ──────────────────────────────────────────── | |
| def _insert_test_data(conn): | |
| """Insert synthetic test data for pair aggregation tests.""" | |
| # 3 genes, 2 cell lines | |
| conn.execute("INSERT INTO genes (gene_id, entrez_id, gene_symbol) VALUES (1, 7157, 'TP53')") | |
| conn.execute("INSERT INTO genes (gene_id, entrez_id, gene_symbol) VALUES (2, 673, 'BRAF')") | |
| conn.execute("INSERT INTO genes (gene_id, entrez_id, gene_symbol) VALUES (3, 2064, 'ERBB2')") | |
| conn.execute("INSERT INTO cell_lines (cell_line_id, model_id) VALUES (1, 'ACH-000001')") | |
| conn.execute("INSERT INTO cell_lines (cell_line_id, model_id) VALUES (2, 'ACH-000002')") | |
| # Screen | |
| conn.execute( | |
| """INSERT INTO ge_screens (screen_id, source_db, depmap_release, screen_type, algorithm) | |
| VALUES (1, 'depmap', '25Q3', 'crispr', 'Chronos')""" | |
| ) | |
| conn.execute( | |
| """INSERT INTO ge_screens (screen_id, source_db, depmap_release, screen_type, algorithm) | |
| VALUES (2, 'demeter2', 'DEMETER2_v6', 'rnai', 'DEMETER2')""" | |
| ) | |
| # Gene 1 (TP53) non-essential in both cell lines (CRISPR) | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, screen_id, gene_effect_score, dependency_probability, | |
| evidence_type, confidence_tier, source_db, source_record_id, extraction_method) | |
| VALUES (1, 1, 1, 0.05, 0.1, 'crispr_nonessential', 'gold', 'depmap', 'r1', 'score_threshold')""" | |
| ) | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, screen_id, gene_effect_score, dependency_probability, | |
| evidence_type, confidence_tier, source_db, source_record_id, extraction_method) | |
| VALUES (1, 2, 1, 0.02, 0.15, 'crispr_nonessential', 'silver', 'depmap', 'r2', 'score_threshold')""" | |
| ) | |
| # Gene 1 (TP53) also non-essential in cell line 1 via RNAi | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, screen_id, gene_effect_score, dependency_probability, | |
| evidence_type, confidence_tier, source_db, source_record_id, extraction_method) | |
| VALUES (1, 1, 2, 0.08, NULL, 'rnai_nonessential', 'bronze', 'demeter2', 'r3', 'score_threshold')""" | |
| ) | |
| # Gene 2 (BRAF) non-essential in cell line 1 only | |
| conn.execute( | |
| """INSERT INTO ge_negative_results | |
| (gene_id, cell_line_id, screen_id, gene_effect_score, dependency_probability, | |
| evidence_type, confidence_tier, source_db, source_record_id, extraction_method) | |
| VALUES (2, 1, 1, -0.3, 0.35, 'crispr_nonessential', 'bronze', 'depmap', 'r4', 'score_threshold')""" | |
| ) | |
| conn.commit() | |
| class TestPairAggregation: | |
| def test_refresh_pair_count(self, conn): | |
| """Should create correct number of aggregated pairs.""" | |
| _insert_test_data(conn) | |
| count = refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| # Gene 1 in cell line 1 (2 sources), Gene 1 in cell line 2 (1 source), | |
| # Gene 2 in cell line 1 (1 source) = 3 pairs | |
| assert count == 3 | |
| def test_multi_source_pair(self, conn): | |
| """Pair with CRISPR + RNAi should have num_sources=2.""" | |
| _insert_test_data(conn) | |
| refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| row = conn.execute( | |
| """SELECT num_screens, num_sources, best_confidence, min_gene_effect, max_gene_effect | |
| FROM gene_cell_pairs WHERE gene_id = 1 AND cell_line_id = 1""" | |
| ).fetchone() | |
| assert row is not None | |
| assert row[0] == 2 # num_screens | |
| assert row[1] == 2 # num_sources (depmap + demeter2) | |
| assert row[2] == "gold" # best_confidence | |
| assert row[3] == pytest.approx(0.05) # min_gene_effect | |
| assert row[4] == pytest.approx(0.08) # max_gene_effect | |
| def test_single_source_pair(self, conn): | |
| """Pair with single source should have num_sources=1.""" | |
| _insert_test_data(conn) | |
| refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| row = conn.execute( | |
| """SELECT num_screens, num_sources, best_confidence | |
| FROM gene_cell_pairs WHERE gene_id = 2 AND cell_line_id = 1""" | |
| ).fetchone() | |
| assert row is not None | |
| assert row[0] == 1 # num_screens | |
| assert row[1] == 1 # num_sources | |
| assert row[2] == "bronze" | |
| def test_gene_degree(self, conn): | |
| """Gene 1 should have degree 2 (non-essential in 2 cell lines).""" | |
| _insert_test_data(conn) | |
| refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| row = conn.execute( | |
| "SELECT gene_degree FROM gene_cell_pairs WHERE gene_id = 1 LIMIT 1" | |
| ).fetchone() | |
| assert row[0] == 2 | |
| def test_cell_line_degree(self, conn): | |
| """Cell line 1 should have degree 2 (2 genes non-essential).""" | |
| _insert_test_data(conn) | |
| refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| row = conn.execute( | |
| "SELECT cell_line_degree FROM gene_cell_pairs WHERE cell_line_id = 1 LIMIT 1" | |
| ).fetchone() | |
| assert row[0] == 2 | |
| def test_mean_gene_effect(self, conn): | |
| """Mean gene effect should be computed correctly.""" | |
| _insert_test_data(conn) | |
| refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| row = conn.execute( | |
| """SELECT mean_gene_effect FROM gene_cell_pairs | |
| WHERE gene_id = 1 AND cell_line_id = 1""" | |
| ).fetchone() | |
| # Average of 0.05 (CRISPR) and 0.08 (RNAi) = 0.065 | |
| assert row[0] == pytest.approx(0.065) | |
| def test_refresh_clears_old_pairs(self, conn): | |
| """Refreshing should delete old pairs and split assignments.""" | |
| _insert_test_data(conn) | |
| refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| # Add a split assignment | |
| conn.execute( | |
| """INSERT INTO ge_split_definitions | |
| (split_name, split_strategy) VALUES ('test_split', 'random')""" | |
| ) | |
| pair_id = conn.execute("SELECT pair_id FROM gene_cell_pairs LIMIT 1").fetchone()[0] | |
| conn.execute( | |
| """INSERT INTO ge_split_assignments (pair_id, split_id, fold) | |
| VALUES (?, 1, 'train')""", | |
| (pair_id,), | |
| ) | |
| conn.commit() | |
| # Refresh again | |
| count = refresh_all_ge_pairs(conn) | |
| conn.commit() | |
| assert count == 3 | |
| # Split assignments should be cleared | |
| sa_count = conn.execute( | |
| "SELECT COUNT(*) FROM ge_split_assignments" | |
| ).fetchone()[0] | |
| assert sa_count == 0 | |
| def test_empty_results(self, conn): | |
| """Refreshing with no results should produce 0 pairs.""" | |
| count = refresh_all_ge_pairs(conn) | |
| assert count == 0 | |
| # ── Index tests ─────────────────────────────────────────────────────── | |
| class TestIndices: | |
| def test_key_indices_exist(self, conn): | |
| """Critical indices should exist for query performance.""" | |
| indices = { | |
| row[1] | |
| for row in conn.execute("PRAGMA index_list('ge_negative_results')").fetchall() | |
| } | |
| assert "idx_ge_nr_gene" in indices | |
| assert "idx_ge_nr_cell_line" in indices | |
| assert "idx_ge_nr_pair" in indices | |
| assert "idx_ge_nr_tier" in indices | |
| assert "idx_ge_nr_unique_source" in indices | |
| def test_cell_line_indices(self, conn): | |
| """Cell line lookup indices should exist.""" | |
| indices = { | |
| row[1] | |
| for row in conn.execute("PRAGMA index_list('cell_lines')").fetchall() | |
| } | |
| assert "idx_cell_lines_ccle" in indices | |
| assert "idx_cell_lines_stripped" in indices | |
| def test_prism_indices(self, conn): | |
| """PRISM bridge indices should exist.""" | |
| indices = { | |
| row[1] | |
| for row in conn.execute("PRAGMA index_list('prism_compounds')").fetchall() | |
| } | |
| assert "idx_prism_inchikey" in indices | |
| assert "idx_prism_chembl" in indices | |