From 84ac477198bba42fd8cee3151badb309e8d5412f Mon Sep 17 00:00:00 2001
From: dbampalikis <dimitris.bampalikis@gmail.com>
Date: Mon, 10 Mar 2025 10:22:05 +0100
Subject: [PATCH] [docs] Add database initial schema

---
 docs/database-schema.dbml | 65 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 65 insertions(+)
 create mode 100644 docs/database-schema.dbml

diff --git a/docs/database-schema.dbml b/docs/database-schema.dbml
new file mode 100644
index 0000000..c4b41dd
--- /dev/null
+++ b/docs/database-schema.dbml
@@ -0,0 +1,65 @@
+// Use DBML to define your database structure
+// Docs: https://dbml.dbdiagram.io/docs
+
+Table compounds {
+  Biocide_name varchar
+  CAS_Number varchar
+  Chemical_class varchar
+}
+
+Table experimental {
+  BacMet_ID varchar
+  Gene_name varchar
+  Code_For varchar
+  Family varchar
+  Proteinaccession_NCBI varchar //Could be foreign key?
+  Nucleotide_accession_ENA_EMBL varchar //Could be foreign key?
+  Protein_accession_Uniprot varchar //Could be foreign key?
+  Organism varchar //Could be foreign key?
+  Location varchar //Could be foreign key?
+  Type_of_compounds	varchar //Could be foreign key?
+  Compound varchar // Foreign key to compounds table
+  Description varchar
+  Length_aa	varchar
+  Reference varchar
+  Number_of_homologues int
+}
+
+Table groups {
+  Unique_ID varchar //The second part seems to be the Proteinaccession_NCBI
+  Sequence varchar
+  Matching_IDs varchar //This is definitely not a varchar
+
+}
+
+Table homologues {
+  query varchar
+  target varchar
+  fident float
+  alnlen int
+  mismatch int
+  gapopen int
+  qstart int
+  qend int
+  qlen int
+  tstart int
+  tend int
+  tlen int
+  evalue int //Very big number
+  bits int
+  prob int
+  lddt float
+  alntmscore float
+  rmsd float
+}
+
+Table pdb {
+  Proteinaccession_NCBI varchar //This is the filename
+  Location varchar // I am guessing we will store these files somewhere
+}
+
+Ref experimental: experimental.Compound < compounds.Biocide_name
+
+Ref homologues: homologues.target < groups.Matching_IDs
+
+Ref experimental: pdb.Proteinaccession_NCBI < experimental.Proteinaccession_NCBI
\ No newline at end of file