tweak vcf schema

Author: mlin

README.md

@@ -1,4 +1,4 @@
-# Genomics Extensions for SQLite
+# Genomics Extension for SQLite
 
 ![build](https://github.com/mlin/GenomicSQLite/workflows/build/badge.svg?branch=main)
 

loaders/vcf_into_sqlite.cc

@@ -49,7 +49,7 @@ string schemaDDL(const string &table_prefix, vector<map<string, string>> &info_h
 
     OStringStream out;
     out << "CREATE TABLE " << table_prefix
-        << "variants (rowid INTEGER NOT NULL PRIMARY KEY, rid INTEGER NOT NULL REFERENCES __gri_refseq(gri_rid), "
+        << "variants (variant_rowid INTEGER NOT NULL PRIMARY KEY, rid INTEGER NOT NULL REFERENCES __gri_refseq(gri_rid), "
            "POS INTEGER NOT NULL, rlen INTEGER NOT NULL, ID_jsarray TEXT, REF TEXT NOT NULL, "
            "ALT_jsarray TEXT, QUAL REAL, FILTER_jsarray";
 
@@ -82,11 +82,11 @@ string schemaDDL(const string &table_prefix, vector<map<string, string>> &info_h
     if (!format_hrecs.empty()) {
         // TODO: include metadata from SAMPLE header lines
         out << ";\nCREATE TABLE " << table_prefix
-            << "samples (rowid INTEGER NOT NULL PRIMARY KEY, id TEXT NOT NULL)";
+            << "samples (sample_id INTEGER NOT NULL PRIMARY KEY, sample_name TEXT NOT NULL)";
         out << ";\nCREATE TABLE " << table_prefix
-            << "genotypes (variant INTEGER NOT NULL REFERENCES " << table_prefix
-            << "variants(rowid), sample INTEGER NOT NULL REFERENCES " << table_prefix
-            << "samples(rowid)";
+            << "genotypes (variant_rowid INTEGER NOT NULL REFERENCES " << table_prefix
+            << "variants(variant_rowid), sample_id INTEGER NOT NULL REFERENCES " << table_prefix
+            << "samples(sample_id)";
 
         // FORMAT columns
         for (auto &hrec : format_hrecs) {
@@ -120,7 +120,7 @@ string schemaDDL(const string &table_prefix, vector<map<string, string>> &info_h
                 }
             }
         }
-        out << "\n, PRIMARY KEY (variant, sample)) WITHOUT ROWID";
+        out << "\n, PRIMARY KEY (variant_rowid, sample_id)) WITHOUT ROWID";
     }
 
     return string(out.Get());
@@ -156,7 +156,8 @@ unique_ptr<SQLite::Statement> prepare_insert_genotype(const string &table_prefix
 }
 
 void insert_samples(bcf_hdr_t *hdr, const string &table_prefix, SQLite::Database &db) {
-    SQLite::Statement stmt(db, "INSERT INTO " + table_prefix + "samples(rowid,id) VALUES(?,?)");
+    SQLite::Statement stmt(db, "INSERT INTO " + table_prefix +
+                                   "samples(sample_id,sample_name) VALUES(?,?)");
 
     for (sqlite3_int64 i = 0; i < bcf_hdr_nsamples(hdr); ++i) {
         stmt.bind(1, i);

test/test_bam.py

@@ -21,12 +21,7 @@ def test_bam(tmp_path):
     bamfile = os.path.join(HERE, "data/NA12878.chr21:20000000-22500000.bam")
     dbfile = str(tmp_path / "test.bam.sqlite")
 
-    sam_into_sqlite(
-        bamfile,
-        str(dbfile),
-        "--table-prefix",
-        "NA12878_"
-    )
+    sam_into_sqlite(bamfile, str(dbfile), "--table-prefix", "NA12878_")
 
     con = genomicsqlite.connect(dbfile, read_only=True)
 
@@ -36,10 +31,14 @@ def test_bam(tmp_path):
     count = next(con.execute("SELECT COUNT(DISTINCT qname) FROM NA12878_reads_seqs"))[0]
     assert count == 299205
 
-    mq_hist = dict(con.execute("""
+    mq_hist = dict(
+        con.execute(
+            """
         SELECT mq, COUNT(*) as count FROM
             (SELECT ifnull(json_extract(tags_json, '$.MQ'),0) AS mq
              FROM NA12878_reads NATURAL JOIN NA12878_reads_tags WHERE (flag & 3840) = 0)
         GROUP BY mq ORDER BY mq DESC
-    """))
-    assert (mq_hist[0],mq_hist[60]) == (2734,520522)
+    """
+        )
+    )
+    assert (mq_hist[0], mq_hist[60]) == (2734, 520522)

test/test_vcf.py

@@ -30,7 +30,7 @@ def test_gnomad_sites_small(tmp_path):
     con = genomicsqlite.connect(dbfile, read_only=True)
 
     query = (
-        "SELECT gnomad_variants._rowid_, id_jsarray FROM (SELECT gri_rid FROM __gri_refseq WHERE gri_refseq_name=?1) AS chrom, gnomad_variants WHERE gnomad_variants._rowid_ IN"
+        "SELECT variant_rowid, id_jsarray FROM (SELECT gri_rid FROM __gri_refseq WHERE gri_refseq_name=?1) AS chrom, gnomad_variants WHERE variant_rowid IN"
         + genomicsqlite.genomic_range_rowids_sql(con, "gnomad_variants", "chrom.gri_rid")
     )
     rs671 = ("chr12", 111803912, 111804012)
@@ -57,9 +57,7 @@ def test_gnomad_sites_small(tmp_path):
     results_rowids = set(vt[0] for vt in results)
     assert next(vt for vt in results if vt[1] and "rs671" in vt[1])
 
-    control = (
-        "SELECT gnomad_variants.rowid FROM gnomad_variants WHERE NOT ((pos+rlen) < ? OR pos > ?)"
-    )
+    control = "SELECT variant_rowid FROM gnomad_variants WHERE NOT ((pos+rlen) < ? OR pos > ?)"
     control_rowids = set(vt[0] for vt in con.execute(control, (rs671[1], rs671[2])))
     assert len(control_rowids) == 22
     assert results_rowids == control_rowids