From b5cda5693bb2742e663effd73afb8c6669fda248 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 15:07:58 +0200
Subject: [PATCH 01/11] Move to psql

---
 Gemfile             |  4 ++--
 Gemfile.lock        |  6 +++++-
 config/database.yml | 11 ++++++-----
 3 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/Gemfile b/Gemfile
index 757c4fa..e80f2e9 100644
--- a/Gemfile
+++ b/Gemfile
@@ -6,8 +6,8 @@ ruby "3.1.2"
 # Bundle edge Rails instead: gem "rails", github: "rails/rails", branch: "main"
 gem "rails", "~> 7.0.3", ">= 7.0.3.1"
 
-# Use mysql as the database for Active Record
-gem "mysql2", "~> 0.5", "> 0.5.5"
+# Use Postgres as the database for Active Record
+gem "pg", "~> 1.5.6"
 
 # Use the Puma web server [https://github.com/puma/puma]
 gem "puma", "~> 5.0"
diff --git a/Gemfile.lock b/Gemfile.lock
index c1fe9b1..902feb1 100644
--- a/Gemfile.lock
+++ b/Gemfile.lock
@@ -119,7 +119,10 @@ GEM
     minitest (5.18.0)
     msgpack (1.5.6)
     multi_json (1.15.0)
+<<<<<<< HEAD
     mysql2 (0.5.6)
+=======
+>>>>>>> b4f1356 (Move to psql)
     net-imap (0.2.3)
       digest
       net-protocol
@@ -150,6 +153,7 @@ GEM
     parallel (1.22.1)
     parser (3.1.2.1)
       ast (~> 2.4.1)
+    pg (1.5.6)
     public_suffix (5.0.0)
     puma (5.6.5)
       nio4r (~> 2.0)
@@ -253,9 +257,9 @@ DEPENDENCIES
   debug
   jbuilder
   multi_json (~> 1.15.0)
-  mysql2 (~> 0.5, > 0.5.5)
   octokit (~> 5.4)
   oj (~> 3.13)
+  pg (~> 1.5.6)
   puma (~> 5.0)
   rails (~> 7.0.3, >= 7.0.3.1)
   rails-controller-testing (~> 1.0)
diff --git a/config/database.yml b/config/database.yml
index 06f9ee6..85ef32e 100644
--- a/config/database.yml
+++ b/config/database.yml
@@ -1,16 +1,17 @@
 default: &default
-  adapter: mysql2
+  adapter: postgresql
   username: root
-  password: unipept
+  password: root
   pool: <%= ENV.fetch("RAILS_MAX_THREADS") { 5 } %>
 
 development:
   <<: *default
-  username: unipept
-  database: unipept
+  username: root
+  database: postgres
   # setup local port forwarding for this to work
   host: 127.0.0.1
-  port: 3306
+  port: 5432
+  schema_search_path: "unipept"
 
 # Warning: The database defined as "test" will be erased and
 # re-generated from your development database when you run "rake".

From 9b052442d6fe1bc9d71035d08642c3298300d646 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 15:15:45 +0200
Subject: [PATCH 02/11] Fix conflict

---
 Gemfile.lock | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/Gemfile.lock b/Gemfile.lock
index 902feb1..95cbf2c 100644
--- a/Gemfile.lock
+++ b/Gemfile.lock
@@ -119,10 +119,6 @@ GEM
     minitest (5.18.0)
     msgpack (1.5.6)
     multi_json (1.15.0)
-<<<<<<< HEAD
-    mysql2 (0.5.6)
-=======
->>>>>>> b4f1356 (Move to psql)
     net-imap (0.2.3)
       digest
       net-protocol

From 0f825cc7bd413f6699cb3f93e3d51f485208eba1 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 15:16:59 +0200
Subject: [PATCH 03/11] Update credentials

---
 config/database.yml | 2 --
 1 file changed, 2 deletions(-)

diff --git a/config/database.yml b/config/database.yml
index 85ef32e..e494dcf 100644
--- a/config/database.yml
+++ b/config/database.yml
@@ -18,9 +18,7 @@ development:
 # Do not set this db to the same as development or production.
 test:
   <<: *default
-  database: unipept_test
   url: <%= ENV['TEST_DATABASE_URL'] || ENV['DATABASE_URL'] %>
 
 production:
   <<: *default
-  database: unipept

From c0f0db83dd12eb4ad5cd35eea64b7f7a9ca85804 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 15:29:40 +0200
Subject: [PATCH 04/11] Update CI

---
 .github/workflows/test.yml | 16 ++++++++--------
 config/database.yml        |  9 +++++----
 2 files changed, 13 insertions(+), 12 deletions(-)

diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index 3eca0da..20a2342 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -6,18 +6,18 @@ jobs:
   test:
     env:
       RAILS_ENV: "test"
-      TEST_DATABASE_URL: "mysql2://root:unipept@127.0.0.1:3306/unipept_test"
+      TEST_DATABASE_URL: "pg://unipept:unipept@127.0.0.1:5432/unipept_test"
     runs-on: ubuntu-latest
     services:
-      mysql:
-        image: mariadb:10.9
+      psql:
+        image: postgres:16.2
         env:
-          MYSQL_DATABASE: "unipept_test"
-          MYSQL_ROOT_PASSWORD: "unipept"
-          MYSQL_HOST: "localhost"
+          POSTGRES_DB: "unipept_test"
+          POSTGRES_PASSWORD: "unipept"
+          POSTGRES_USER: "unipept"
         ports:
-          - 3306:3306
-        options: --health-cmd "mysqladmin ping -h localhost" --health-interval 10s --health-timeout 5s --health-retries 5
+          - 5432:5432
+        options: --health-cmd "pg_isready" --health-interval 10s --health-timeout 5s --health-retries 5
     steps:
       - uses: actions/checkout@v4
       - name: Use ruby from .ruby-version
diff --git a/config/database.yml b/config/database.yml
index e494dcf..4994e36 100644
--- a/config/database.yml
+++ b/config/database.yml
@@ -1,13 +1,12 @@
 default: &default
   adapter: postgresql
-  username: root
-  password: root
+  username: unipept
+  password: unipept
   pool: <%= ENV.fetch("RAILS_MAX_THREADS") { 5 } %>
 
 development:
   <<: *default
-  username: root
-  database: postgres
+  database: unipept
   # setup local port forwarding for this to work
   host: 127.0.0.1
   port: 5432
@@ -18,7 +17,9 @@ development:
 # Do not set this db to the same as development or production.
 test:
   <<: *default
+  database: unipept_test
   url: <%= ENV['TEST_DATABASE_URL'] || ENV['DATABASE_URL'] %>
 
 production:
   <<: *default
+  database: unipept

From 3e1e63a482298be858a18d104ce30425ab75278f Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 15:31:13 +0200
Subject: [PATCH 05/11] Update CI

---
 .github/workflows/test.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index 20a2342..9e56507 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -6,7 +6,7 @@ jobs:
   test:
     env:
       RAILS_ENV: "test"
-      TEST_DATABASE_URL: "pg://unipept:unipept@127.0.0.1:5432/unipept_test"
+      TEST_DATABASE_URL: "postgresql://unipept:unipept@127.0.0.1:5432/unipept_test"
     runs-on: ubuntu-latest
     services:
       psql:

From 7f98aec2eb4d31e976d8a1730cb5b63fd53d3df2 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 15:38:02 +0200
Subject: [PATCH 06/11] Migrate tests

---
 db/schema.rb | 40 ++++++++++++++++++++--------------------
 1 file changed, 20 insertions(+), 20 deletions(-)

diff --git a/db/schema.rb b/db/schema.rb
index 9bc0487..73831b0 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -12,7 +12,7 @@
 
 ActiveRecord::Schema.define(version: 0) do
 
-  create_table "dataset_items", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+  create_table "dataset_items", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "dataset_id", unsigned: true
     t.string "name", limit: 160
     t.text "data", limit: 16777215, null: false, collation: "ascii_general_ci"
@@ -20,52 +20,52 @@
     t.index ["dataset_id"], name: "fk_dataset_items_datasets"
   end
 
-  create_table "datasets", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+  create_table "datasets", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "environment", limit: 160
     t.string "reference", limit: 500
     t.string "url", limit: 200
     t.string "project_website", limit: 200
   end
 
-  create_table "ec_cross_references", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "ec_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "uniprot_entry_id", null: false, unsigned: true
     t.string "ec_number_code", limit: 15, null: false
   end
 
-  create_table "ec_numbers", id: :integer, limit: 2, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+  create_table "ec_numbers", id: :integer, limit: 2, unsigned: true, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.string "name", limit: 155, null: false
   end
 
-  create_table "embl_cross_references", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "embl_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "uniprot_entry_id", null: false, unsigned: true
     t.string "protein_id", limit: 25
     t.string "sequence_id", limit: 25
   end
 
-  create_table "go_cross_references", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "go_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "uniprot_entry_id", null: false, unsigned: true
     t.string "go_term_code", limit: 15, null: false
   end
 
-  create_table "go_terms", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=latin1", force: :cascade do |t|
+  create_table "go_terms", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.string "namespace", limit: 18, null: false
     t.string "name", limit: 200, null: false
   end
 
-  create_table "interpro_cross_references", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "interpro_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "uniprot_entry_id", null: false, unsigned: true
     t.string "interpro_entry_code", limit: 9, null: false
   end
 
-  create_table "interpro_entries", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "interpro_entries", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "code", limit: 9, null: false
     t.string "category", limit: 32, null: false
     t.string "name", limit: 160, null: false
   end
 
-  create_table "lineages", primary_key: "taxon_id", id: :integer, limit: 3, unsigned: true, default: nil, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "lineages", primary_key: "taxon_id", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
     t.integer "superkingdom", limit: 3
     t.integer "kingdom", limit: 3
     t.integer "subkingdom", limit: 3
@@ -95,28 +95,28 @@
     t.integer "forma", limit: 3
   end
 
-  create_table "peptides", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "peptides", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "sequence_id", null: false, unsigned: true
     t.integer "original_sequence_id", null: false, unsigned: true
     t.integer "uniprot_entry_id", null: false, unsigned: true
   end
 
-  create_table "posts", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+  create_table "posts", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "title", limit: 100, null: false
     t.text "content", null: false
     t.date "date", null: false
   end
 
-  create_table "proteome_caches", primary_key: "proteome_id", id: :integer, limit: 3, unsigned: true, default: nil, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "proteome_caches", primary_key: "proteome_id", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
     t.text "json_sequences", limit: 16777215, null: false
   end
 
-  create_table "proteome_cross_references", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=latin1", force: :cascade do |t|
+  create_table "proteome_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "uniprot_entry_id", null: false, unsigned: true
     t.integer "proteome_id", limit: 3, null: false, unsigned: true
   end
 
-  create_table "proteomes", id: :integer, limit: 3, unsigned: true, default: nil, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+  create_table "proteomes", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
     t.string "proteome_accession_number", limit: 12, null: false
     t.string "proteome_name", limit: 145, null: false
     t.integer "taxon_id", limit: 3, unsigned: true
@@ -127,13 +127,13 @@
     t.string "name", limit: 225
   end
 
-  create_table "refseq_cross_references", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "refseq_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "uniprot_entry_id", null: false, unsigned: true
     t.string "protein_id", limit: 25
     t.string "sequence_id", limit: 25
   end
 
-  create_table "sequences", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii ROW_FORMAT=COMPRESSED KEY_BLOCK_SIZE=16", force: :cascade do |t|
+  create_table "sequences", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "sequence", limit: 50, null: false
     t.integer "lca", limit: 3, unsigned: true
     t.integer "lca_il", limit: 3, unsigned: true
@@ -141,14 +141,14 @@
     t.binary "fa_il", limit: 16777215
   end
 
-  create_table "taxons", id: :integer, limit: 3, unsigned: true, default: nil, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+  create_table "taxons", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
     t.string "name", limit: 120, null: false
     t.string "rank", limit: 16
     t.integer "parent_id", limit: 3, unsigned: true
     t.binary "valid_taxon", limit: 1, default: 0b1, null: false
   end
 
-  create_table "uniprot_entries", id: :integer, unsigned: true, options: "ENGINE=InnoDB DEFAULT CHARSET=ascii", force: :cascade do |t|
+  create_table "uniprot_entries", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "uniprot_accession_number", limit: 10, null: false, collation: "latin1_swedish_ci"
     t.integer "version", limit: 2, null: false, unsigned: true
     t.integer "taxon_id", limit: 3, null: false, unsigned: true
@@ -157,7 +157,7 @@
     t.text "protein", null: false
   end
 
-  create_table "users", id: :integer, options: "ENGINE=InnoDB DEFAULT CHARSET=latin1", force: :cascade do |t|
+  create_table "users", id: :integer, force: :cascade do |t|
     t.string "username", limit: 8, null: false
     t.integer "admin", limit: 1, default: 0, null: false
   end

From 71e0339aeb11f17745b262075374021ca0ef3cef Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 16:00:15 +0200
Subject: [PATCH 07/11] Re-generate schema

---
 db/schema.rb | 205 +++++++++++++++++++++++----------------------------
 1 file changed, 91 insertions(+), 114 deletions(-)

diff --git a/db/schema.rb b/db/schema.rb
index 73831b0..29a7054 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -2,164 +2,141 @@
 # of editing this file, please use the migrations feature of Active Record to
 # incrementally modify your database, and then regenerate this schema definition.
 #
-# Note that this schema.rb definition is the authoritative source for your
-# database schema. If you need to create the application database on another
-# system, you should be using db:schema:load, not running all the migrations
-# from scratch. The latter is a flawed and unsustainable approach (the more migrations
-# you'll amass, the slower it'll run and the greater likelihood for issues).
+# This file is the source Rails uses to define your schema when running `bin/rails
+# db:schema:load`. When creating a new database, `bin/rails db:schema:load` tends to
+# be faster and is potentially less error prone than running all of your
+# migrations from scratch. Old migrations may fail to apply correctly if those
+# migrations use external dependencies or application code.
 #
 # It's strongly recommended that you check this file into your version control system.
 
-ActiveRecord::Schema.define(version: 0) do
+ActiveRecord::Schema[7.0].define(version: 0) do
+  # These are extensions that must be enabled in order to support this database
+  enable_extension "plpgsql"
 
-  create_table "dataset_items", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "dataset_id", unsigned: true
+  # Custom types defined in this database.
+  # Note that some types may not work with other database engines. Be careful if changing database.
+  create_enum "db_type", ["swissprot", "trembl"]
+  create_enum "go_namespace", ["biological process", "molecular function", "cellular component"]
+  create_enum "rank_type", ["no rank", "superkingdom", "kingdom", "subkingdom", "superphylum", "phylum", "subphylum", "superclass", "class", "subclass", "superorder", "order", "suborder", "infraorder", "superfamily", "family", "subfamily", "tribe", "subtribe", "genus", "subgenus", "species group", "species subgroup", "species", "subspecies", "strain", "varietas", "forma"]
+
+  create_table "dataset_items", id: :integer, default: nil, force: :cascade do |t|
+    t.bigint "dataset_id"
     t.string "name", limit: 160
-    t.text "data", limit: 16777215, null: false, collation: "ascii_general_ci"
+    t.text "data", null: false
     t.integer "order"
-    t.index ["dataset_id"], name: "fk_dataset_items_datasets"
   end
 
-  create_table "datasets", id: :integer, unsigned: true, force: :cascade do |t|
+  create_table "datasets", id: :integer, default: nil, force: :cascade do |t|
     t.string "environment", limit: 160
     t.string "reference", limit: 500
     t.string "url", limit: 200
     t.string "project_website", limit: 200
   end
 
-  create_table "ec_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "uniprot_entry_id", null: false, unsigned: true
+  create_table "ec_cross_references", id: :bigint, default: nil, force: :cascade do |t|
+    t.bigint "uniprot_entry_id", null: false
     t.string "ec_number_code", limit: 15, null: false
+    t.index ["uniprot_entry_id"], name: "idx_ec_cross_references_uniprot_entry_id"
   end
 
-  create_table "ec_numbers", id: :integer, limit: 2, unsigned: true, force: :cascade do |t|
+  create_table "ec_numbers", id: :integer, default: nil, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.string "name", limit: 155, null: false
+    t.index ["code"], name: "idx_ec_numbers_code"
   end
 
-  create_table "embl_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "uniprot_entry_id", null: false, unsigned: true
-    t.string "protein_id", limit: 25
-    t.string "sequence_id", limit: 25
-  end
-
-  create_table "go_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "uniprot_entry_id", null: false, unsigned: true
+  create_table "go_cross_references", id: :bigint, default: nil, force: :cascade do |t|
+    t.bigint "uniprot_entry_id", null: false
     t.string "go_term_code", limit: 15, null: false
+    t.index ["uniprot_entry_id"], name: "idx_go_cross_references_uniprot_entry_id"
   end
 
-  create_table "go_terms", id: :integer, unsigned: true, force: :cascade do |t|
+  create_table "go_terms", id: :integer, default: nil, force: :cascade do |t|
     t.string "code", limit: 15, null: false
-    t.string "namespace", limit: 18, null: false
+    t.enum "namespace", null: false, enum_type: "go_namespace"
     t.string "name", limit: 200, null: false
+    t.index ["code"], name: "idx_go_terms_code"
   end
 
-  create_table "interpro_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "uniprot_entry_id", null: false, unsigned: true
+  create_table "interpro_cross_references", id: :bigint, default: nil, force: :cascade do |t|
+    t.bigint "uniprot_entry_id", null: false
     t.string "interpro_entry_code", limit: 9, null: false
+    t.index ["uniprot_entry_id"], name: "idx_interpro_cross_references_uniprot_entry_id"
   end
 
-  create_table "interpro_entries", id: :integer, unsigned: true, force: :cascade do |t|
+  create_table "interpro_entries", id: :integer, default: nil, force: :cascade do |t|
     t.string "code", limit: 9, null: false
     t.string "category", limit: 32, null: false
     t.string "name", limit: 160, null: false
   end
 
-  create_table "lineages", primary_key: "taxon_id", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
-    t.integer "superkingdom", limit: 3
-    t.integer "kingdom", limit: 3
-    t.integer "subkingdom", limit: 3
-    t.integer "superphylum", limit: 3
-    t.integer "phylum", limit: 3
-    t.integer "subphylum", limit: 3
-    t.integer "superclass", limit: 3
-    t.integer "class", limit: 3
-    t.integer "subclass", limit: 3
-    t.integer "superorder", limit: 3
-    t.integer "order", limit: 3
-    t.integer "suborder", limit: 3
-    t.integer "infraorder", limit: 3
-    t.integer "superfamily", limit: 3
-    t.integer "family", limit: 3
-    t.integer "subfamily", limit: 3
-    t.integer "tribe", limit: 3
-    t.integer "subtribe", limit: 3
-    t.integer "genus", limit: 3
-    t.integer "subgenus", limit: 3
-    t.integer "species_group", limit: 3
-    t.integer "species_subgroup", limit: 3
-    t.integer "species", limit: 3
-    t.integer "subspecies", limit: 3
-    t.integer "strain", limit: 3
-    t.integer "varietas", limit: 3
-    t.integer "forma", limit: 3
-  end
-
-  create_table "peptides", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "sequence_id", null: false, unsigned: true
-    t.integer "original_sequence_id", null: false, unsigned: true
-    t.integer "uniprot_entry_id", null: false, unsigned: true
-  end
-
-  create_table "posts", id: :integer, unsigned: true, force: :cascade do |t|
-    t.string "title", limit: 100, null: false
-    t.text "content", null: false
-    t.date "date", null: false
-  end
-
-  create_table "proteome_caches", primary_key: "proteome_id", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
-    t.text "json_sequences", limit: 16777215, null: false
-  end
-
-  create_table "proteome_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "uniprot_entry_id", null: false, unsigned: true
-    t.integer "proteome_id", limit: 3, null: false, unsigned: true
-  end
-
-  create_table "proteomes", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
-    t.string "proteome_accession_number", limit: 12, null: false
-    t.string "proteome_name", limit: 145, null: false
-    t.integer "taxon_id", limit: 3, unsigned: true
-    t.binary "type_strain", limit: 1, default: 0b0, null: false
-    t.binary "reference_proteome", limit: 1, default: 0b0, null: false
-    t.string "strain", limit: 120
-    t.string "assembly", limit: 45
-    t.string "name", limit: 225
-  end
-
-  create_table "refseq_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "uniprot_entry_id", null: false, unsigned: true
-    t.string "protein_id", limit: 25
-    t.string "sequence_id", limit: 25
-  end
-
-  create_table "sequences", id: :integer, unsigned: true, force: :cascade do |t|
+  create_table "lineages", primary_key: "taxon_id", id: :integer, default: nil, force: :cascade do |t|
+    t.integer "superkingdom"
+    t.integer "kingdom"
+    t.integer "subkingdom"
+    t.integer "superphylum"
+    t.integer "phylum"
+    t.integer "subphylum"
+    t.integer "superclass"
+    t.integer "class"
+    t.integer "subclass"
+    t.integer "superorder"
+    t.integer "order"
+    t.integer "suborder"
+    t.integer "infraorder"
+    t.integer "superfamily"
+    t.integer "family"
+    t.integer "subfamily"
+    t.integer "tribe"
+    t.integer "subtribe"
+    t.integer "genus"
+    t.integer "subgenus"
+    t.integer "species_group"
+    t.integer "species_subgroup"
+    t.integer "species"
+    t.integer "subspecies"
+    t.integer "strain"
+    t.integer "varietas"
+    t.integer "forma"
+  end
+
+  create_table "peptides", id: :bigint, default: nil, force: :cascade do |t|
+    t.bigint "sequence_id", null: false
+    t.bigint "original_sequence_id", null: false
+    t.bigint "uniprot_entry_id", null: false
+    t.index ["original_sequence_id"], name: "idx_peptides_original_sequence_id"
+    t.index ["sequence_id"], name: "idx_peptides_sequence_id"
+    t.index ["uniprot_entry_id"], name: "idx_peptides_uniprot_entry_id"
+  end
+
+  create_table "sequences", id: :bigint, default: nil, force: :cascade do |t|
     t.string "sequence", limit: 50, null: false
-    t.integer "lca", limit: 3, unsigned: true
-    t.integer "lca_il", limit: 3, unsigned: true
-    t.binary "fa", limit: 16777215
-    t.binary "fa_il", limit: 16777215
+    t.integer "lca"
+    t.integer "lca_il"
+    t.binary "fa"
+    t.binary "fa_il"
+    t.index ["lca"], name: "idx_sequences_lca"
+    t.index ["lca_il"], name: "idx_sequences_lca_il"
+    t.index ["sequence"], name: "idx_sequences_sequence"
   end
 
-  create_table "taxons", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
+  create_table "taxons", id: :integer, default: nil, force: :cascade do |t|
     t.string "name", limit: 120, null: false
-    t.string "rank", limit: 16
-    t.integer "parent_id", limit: 3, unsigned: true
-    t.binary "valid_taxon", limit: 1, default: 0b1, null: false
+    t.enum "rank", enum_type: "rank_type"
+    t.integer "parent_id"
+    t.integer "valid_taxon", limit: 2, default: 1, null: false
   end
 
-  create_table "uniprot_entries", id: :integer, unsigned: true, force: :cascade do |t|
-    t.string "uniprot_accession_number", limit: 10, null: false, collation: "latin1_swedish_ci"
-    t.integer "version", limit: 2, null: false, unsigned: true
-    t.integer "taxon_id", limit: 3, null: false, unsigned: true
-    t.string "type", limit: 9, null: false
+  create_table "uniprot_entries", id: :integer, default: nil, force: :cascade do |t|
+    t.string "uniprot_accession_number", limit: 10, null: false
+    t.integer "version", null: false
+    t.integer "taxon_id", null: false
+    t.enum "type", null: false, enum_type: "db_type"
     t.string "name", limit: 150, null: false
     t.text "protein", null: false
-  end
-
-  create_table "users", id: :integer, force: :cascade do |t|
-    t.string "username", limit: 8, null: false
-    t.integer "admin", limit: 1, default: 0, null: false
+    t.index ["taxon_id"], name: "idx_uniprot_entries_taxon_id"
+    t.index ["uniprot_accession_number"], name: "idx_uniprot_entries_uniprot_accession_number"
   end
 
   add_foreign_key "dataset_items", "datasets", name: "fk_dataset_items_datasets"

From f6f8608ba27fb5fdfe9e6dca6a30c85468bea400 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 16:11:45 +0200
Subject: [PATCH 08/11] Add missing table

---
 db/schema.rb | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/db/schema.rb b/db/schema.rb
index 29a7054..3acef0d 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -139,5 +139,10 @@
     t.index ["uniprot_accession_number"], name: "idx_uniprot_entries_uniprot_accession_number"
   end
 
+  create_table "users", id: :integer, force: :cascade do |t|
+    t.string "username", limit: 8, null: false
+    t.integer "admin", limit: 1, default: 0, null: false
+  end
+
   add_foreign_key "dataset_items", "datasets", name: "fk_dataset_items_datasets"
 end

From edc40faa471c6876dfb0f0ca1e5571e66918dcf1 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 16:18:20 +0200
Subject: [PATCH 09/11] Add auto increment

---
 db/schema.rb | 26 +++++++++++++-------------
 1 file changed, 13 insertions(+), 13 deletions(-)

diff --git a/db/schema.rb b/db/schema.rb
index 3acef0d..94adb57 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -20,58 +20,58 @@
   create_enum "go_namespace", ["biological process", "molecular function", "cellular component"]
   create_enum "rank_type", ["no rank", "superkingdom", "kingdom", "subkingdom", "superphylum", "phylum", "subphylum", "superclass", "class", "subclass", "superorder", "order", "suborder", "infraorder", "superfamily", "family", "subfamily", "tribe", "subtribe", "genus", "subgenus", "species group", "species subgroup", "species", "subspecies", "strain", "varietas", "forma"]
 
-  create_table "dataset_items", id: :integer, default: nil, force: :cascade do |t|
+  create_table "dataset_items", id: :integer, auto_increment: true, force: :cascade do |t|
     t.bigint "dataset_id"
     t.string "name", limit: 160
     t.text "data", null: false
     t.integer "order"
   end
 
-  create_table "datasets", id: :integer, default: nil, force: :cascade do |t|
+  create_table "datasets", id: :integer, auto_increment: true, force: :cascade do |t|
     t.string "environment", limit: 160
     t.string "reference", limit: 500
     t.string "url", limit: 200
     t.string "project_website", limit: 200
   end
 
-  create_table "ec_cross_references", id: :bigint, default: nil, force: :cascade do |t|
+  create_table "ec_cross_references", id: :bigint, auto_increment: true, force: :cascade do |t|
     t.bigint "uniprot_entry_id", null: false
     t.string "ec_number_code", limit: 15, null: false
     t.index ["uniprot_entry_id"], name: "idx_ec_cross_references_uniprot_entry_id"
   end
 
-  create_table "ec_numbers", id: :integer, default: nil, force: :cascade do |t|
+  create_table "ec_numbers", id: :integer, auto_increment: true, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.string "name", limit: 155, null: false
     t.index ["code"], name: "idx_ec_numbers_code"
   end
 
-  create_table "go_cross_references", id: :bigint, default: nil, force: :cascade do |t|
+  create_table "go_cross_references", id: :bigint, auto_increment: true, force: :cascade do |t|
     t.bigint "uniprot_entry_id", null: false
     t.string "go_term_code", limit: 15, null: false
     t.index ["uniprot_entry_id"], name: "idx_go_cross_references_uniprot_entry_id"
   end
 
-  create_table "go_terms", id: :integer, default: nil, force: :cascade do |t|
+  create_table "go_terms", id: :integer, auto_increment: true, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.enum "namespace", null: false, enum_type: "go_namespace"
     t.string "name", limit: 200, null: false
     t.index ["code"], name: "idx_go_terms_code"
   end
 
-  create_table "interpro_cross_references", id: :bigint, default: nil, force: :cascade do |t|
+  create_table "interpro_cross_references", id: :bigint, auto_increment: true, force: :cascade do |t|
     t.bigint "uniprot_entry_id", null: false
     t.string "interpro_entry_code", limit: 9, null: false
     t.index ["uniprot_entry_id"], name: "idx_interpro_cross_references_uniprot_entry_id"
   end
 
-  create_table "interpro_entries", id: :integer, default: nil, force: :cascade do |t|
+  create_table "interpro_entries", id: :integer, auto_increment: true, force: :cascade do |t|
     t.string "code", limit: 9, null: false
     t.string "category", limit: 32, null: false
     t.string "name", limit: 160, null: false
   end
 
-  create_table "lineages", primary_key: "taxon_id", id: :integer, default: nil, force: :cascade do |t|
+  create_table "lineages", primary_key: "taxon_id", id: :integer, auto_increment: true, force: :cascade do |t|
     t.integer "superkingdom"
     t.integer "kingdom"
     t.integer "subkingdom"
@@ -101,7 +101,7 @@
     t.integer "forma"
   end
 
-  create_table "peptides", id: :bigint, default: nil, force: :cascade do |t|
+  create_table "peptides", id: :bigint, auto_increment: true, force: :cascade do |t|
     t.bigint "sequence_id", null: false
     t.bigint "original_sequence_id", null: false
     t.bigint "uniprot_entry_id", null: false
@@ -110,7 +110,7 @@
     t.index ["uniprot_entry_id"], name: "idx_peptides_uniprot_entry_id"
   end
 
-  create_table "sequences", id: :bigint, default: nil, force: :cascade do |t|
+  create_table "sequences", id: :bigint, auto_increment: true, force: :cascade do |t|
     t.string "sequence", limit: 50, null: false
     t.integer "lca"
     t.integer "lca_il"
@@ -121,14 +121,14 @@
     t.index ["sequence"], name: "idx_sequences_sequence"
   end
 
-  create_table "taxons", id: :integer, default: nil, force: :cascade do |t|
+  create_table "taxons", id: :integer, auto_increment: true, force: :cascade do |t|
     t.string "name", limit: 120, null: false
     t.enum "rank", enum_type: "rank_type"
     t.integer "parent_id"
     t.integer "valid_taxon", limit: 2, default: 1, null: false
   end
 
-  create_table "uniprot_entries", id: :integer, default: nil, force: :cascade do |t|
+  create_table "uniprot_entries", id: :integer, auto_increment: true, force: :cascade do |t|
     t.string "uniprot_accession_number", limit: 10, null: false
     t.integer "version", null: false
     t.integer "taxon_id", null: false

From 99f5fb752194bc8235e405305c5790d94e3e93db Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 16:20:18 +0200
Subject: [PATCH 10/11] Remove auto increment

---
 db/schema.rb | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

diff --git a/db/schema.rb b/db/schema.rb
index 94adb57..44ecd03 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -20,58 +20,58 @@
   create_enum "go_namespace", ["biological process", "molecular function", "cellular component"]
   create_enum "rank_type", ["no rank", "superkingdom", "kingdom", "subkingdom", "superphylum", "phylum", "subphylum", "superclass", "class", "subclass", "superorder", "order", "suborder", "infraorder", "superfamily", "family", "subfamily", "tribe", "subtribe", "genus", "subgenus", "species group", "species subgroup", "species", "subspecies", "strain", "varietas", "forma"]
 
-  create_table "dataset_items", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "dataset_items", id: :integer, unsigned: true, force: :cascade do |t|
     t.bigint "dataset_id"
     t.string "name", limit: 160
     t.text "data", null: false
     t.integer "order"
   end
 
-  create_table "datasets", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "datasets", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "environment", limit: 160
     t.string "reference", limit: 500
     t.string "url", limit: 200
     t.string "project_website", limit: 200
   end
 
-  create_table "ec_cross_references", id: :bigint, auto_increment: true, force: :cascade do |t|
+  create_table "ec_cross_references", id: :bigint, force: :cascade do |t|
     t.bigint "uniprot_entry_id", null: false
     t.string "ec_number_code", limit: 15, null: false
     t.index ["uniprot_entry_id"], name: "idx_ec_cross_references_uniprot_entry_id"
   end
 
-  create_table "ec_numbers", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "ec_numbers", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.string "name", limit: 155, null: false
     t.index ["code"], name: "idx_ec_numbers_code"
   end
 
-  create_table "go_cross_references", id: :bigint, auto_increment: true, force: :cascade do |t|
+  create_table "go_cross_references", id: :bigint, force: :cascade do |t|
     t.bigint "uniprot_entry_id", null: false
     t.string "go_term_code", limit: 15, null: false
     t.index ["uniprot_entry_id"], name: "idx_go_cross_references_uniprot_entry_id"
   end
 
-  create_table "go_terms", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "go_terms", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "code", limit: 15, null: false
     t.enum "namespace", null: false, enum_type: "go_namespace"
     t.string "name", limit: 200, null: false
     t.index ["code"], name: "idx_go_terms_code"
   end
 
-  create_table "interpro_cross_references", id: :bigint, auto_increment: true, force: :cascade do |t|
+  create_table "interpro_cross_references", id: :bigint, force: :cascade do |t|
     t.bigint "uniprot_entry_id", null: false
     t.string "interpro_entry_code", limit: 9, null: false
     t.index ["uniprot_entry_id"], name: "idx_interpro_cross_references_uniprot_entry_id"
   end
 
-  create_table "interpro_entries", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "interpro_entries", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "code", limit: 9, null: false
     t.string "category", limit: 32, null: false
     t.string "name", limit: 160, null: false
   end
 
-  create_table "lineages", primary_key: "taxon_id", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "lineages", primary_key: "taxon_id", id: :integer, unsigned: true, force: :cascade do |t|
     t.integer "superkingdom"
     t.integer "kingdom"
     t.integer "subkingdom"
@@ -101,7 +101,7 @@
     t.integer "forma"
   end
 
-  create_table "peptides", id: :bigint, auto_increment: true, force: :cascade do |t|
+  create_table "peptides", id: :bigint, force: :cascade do |t|
     t.bigint "sequence_id", null: false
     t.bigint "original_sequence_id", null: false
     t.bigint "uniprot_entry_id", null: false
@@ -110,7 +110,7 @@
     t.index ["uniprot_entry_id"], name: "idx_peptides_uniprot_entry_id"
   end
 
-  create_table "sequences", id: :bigint, auto_increment: true, force: :cascade do |t|
+  create_table "sequences", id: :bigint, force: :cascade do |t|
     t.string "sequence", limit: 50, null: false
     t.integer "lca"
     t.integer "lca_il"
@@ -121,14 +121,14 @@
     t.index ["sequence"], name: "idx_sequences_sequence"
   end
 
-  create_table "taxons", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "taxons", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "name", limit: 120, null: false
     t.enum "rank", enum_type: "rank_type"
     t.integer "parent_id"
     t.integer "valid_taxon", limit: 2, default: 1, null: false
   end
 
-  create_table "uniprot_entries", id: :integer, auto_increment: true, force: :cascade do |t|
+  create_table "uniprot_entries", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "uniprot_accession_number", limit: 10, null: false
     t.integer "version", null: false
     t.integer "taxon_id", null: false
@@ -139,7 +139,7 @@
     t.index ["uniprot_accession_number"], name: "idx_uniprot_entries_uniprot_accession_number"
   end
 
-  create_table "users", id: :integer, force: :cascade do |t|
+  create_table "users", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "username", limit: 8, null: false
     t.integer "admin", limit: 1, default: 0, null: false
   end

From 16ce1f652e34ef00e2f80bf8702d473dd74e7cb8 Mon Sep 17 00:00:00 2001
From: stijndcl <declercq.stijn@outlook.com>
Date: Mon, 6 May 2024 16:29:54 +0200
Subject: [PATCH 11/11] Manually add missing things back

---
 db/schema.rb | 128 ++++++++++++++++++++++++++++++++-------------------
 1 file changed, 80 insertions(+), 48 deletions(-)

diff --git a/db/schema.rb b/db/schema.rb
index 44ecd03..c3e4150 100644
--- a/db/schema.rb
+++ b/db/schema.rb
@@ -21,7 +21,7 @@
   create_enum "rank_type", ["no rank", "superkingdom", "kingdom", "subkingdom", "superphylum", "phylum", "subphylum", "superclass", "class", "subclass", "superorder", "order", "suborder", "infraorder", "superfamily", "family", "subfamily", "tribe", "subtribe", "genus", "subgenus", "species group", "species subgroup", "species", "subspecies", "strain", "varietas", "forma"]
 
   create_table "dataset_items", id: :integer, unsigned: true, force: :cascade do |t|
-    t.bigint "dataset_id"
+    t.bigint "dataset_id", unsigned: true
     t.string "name", limit: 160
     t.text "data", null: false
     t.integer "order"
@@ -34,8 +34,8 @@
     t.string "project_website", limit: 200
   end
 
-  create_table "ec_cross_references", id: :bigint, force: :cascade do |t|
-    t.bigint "uniprot_entry_id", null: false
+  create_table "ec_cross_references", id: :bigint, unsigned: true, force: :cascade do |t|
+    t.bigint "uniprot_entry_id", null: false, unsigned: true
     t.string "ec_number_code", limit: 15, null: false
     t.index ["uniprot_entry_id"], name: "idx_ec_cross_references_uniprot_entry_id"
   end
@@ -46,8 +46,8 @@
     t.index ["code"], name: "idx_ec_numbers_code"
   end
 
-  create_table "go_cross_references", id: :bigint, force: :cascade do |t|
-    t.bigint "uniprot_entry_id", null: false
+  create_table "go_cross_references", id: :bigint, unsigned: true, force: :cascade do |t|
+    t.bigint "uniprot_entry_id", null: false, unsigned: true
     t.string "go_term_code", limit: 15, null: false
     t.index ["uniprot_entry_id"], name: "idx_go_cross_references_uniprot_entry_id"
   end
@@ -59,8 +59,8 @@
     t.index ["code"], name: "idx_go_terms_code"
   end
 
-  create_table "interpro_cross_references", id: :bigint, force: :cascade do |t|
-    t.bigint "uniprot_entry_id", null: false
+  create_table "interpro_cross_references", id: :bigint, unsigned: true, force: :cascade do |t|
+    t.bigint "uniprot_entry_id", null: false, unsigned: true
     t.string "interpro_entry_code", limit: 9, null: false
     t.index ["uniprot_entry_id"], name: "idx_interpro_cross_references_uniprot_entry_id"
   end
@@ -72,50 +72,82 @@
   end
 
   create_table "lineages", primary_key: "taxon_id", id: :integer, unsigned: true, force: :cascade do |t|
-    t.integer "superkingdom"
-    t.integer "kingdom"
-    t.integer "subkingdom"
-    t.integer "superphylum"
-    t.integer "phylum"
-    t.integer "subphylum"
-    t.integer "superclass"
-    t.integer "class"
-    t.integer "subclass"
-    t.integer "superorder"
-    t.integer "order"
-    t.integer "suborder"
-    t.integer "infraorder"
-    t.integer "superfamily"
-    t.integer "family"
-    t.integer "subfamily"
-    t.integer "tribe"
-    t.integer "subtribe"
-    t.integer "genus"
-    t.integer "subgenus"
-    t.integer "species_group"
-    t.integer "species_subgroup"
-    t.integer "species"
-    t.integer "subspecies"
-    t.integer "strain"
-    t.integer "varietas"
-    t.integer "forma"
-  end
-
-  create_table "peptides", id: :bigint, force: :cascade do |t|
-    t.bigint "sequence_id", null: false
-    t.bigint "original_sequence_id", null: false
-    t.bigint "uniprot_entry_id", null: false
+    t.integer "superkingdom", limit: 3
+    t.integer "kingdom", limit: 3
+    t.integer "subkingdom", limit: 3
+    t.integer "superphylum", limit: 3
+    t.integer "phylum", limit: 3
+    t.integer "subphylum", limit: 3
+    t.integer "superclass", limit: 3
+    t.integer "class", limit: 3
+    t.integer "subclass", limit: 3
+    t.integer "superorder", limit: 3
+    t.integer "order", limit: 3
+    t.integer "suborder", limit: 3
+    t.integer "infraorder", limit: 3
+    t.integer "superfamily", limit: 3
+    t.integer "family", limit: 3
+    t.integer "subfamily", limit: 3
+    t.integer "tribe", limit: 3
+    t.integer "subtribe", limit: 3
+    t.integer "genus", limit: 3
+    t.integer "subgenus", limit: 3
+    t.integer "species_group", limit: 3
+    t.integer "species_subgroup", limit: 3
+    t.integer "species", limit: 3
+    t.integer "subspecies", limit: 3
+    t.integer "strain", limit: 3
+    t.integer "varietas", limit: 3
+    t.integer "forma", limit: 3
+  end
+
+  create_table "peptides", id: :bigint, unsigned: true, force: :cascade do |t|
+    t.bigint "sequence_id", null: false, unsigned: true
+    t.bigint "original_sequence_id", null: false, unsigned: true
+    t.bigint "uniprot_entry_id", null: false, unsigned: true
     t.index ["original_sequence_id"], name: "idx_peptides_original_sequence_id"
     t.index ["sequence_id"], name: "idx_peptides_sequence_id"
     t.index ["uniprot_entry_id"], name: "idx_peptides_uniprot_entry_id"
   end
 
-  create_table "sequences", id: :bigint, force: :cascade do |t|
+  create_table "posts", id: :integer, unsigned: true, force: :cascade do |t|
+    t.string "title", limit: 100, null: false
+    t.text "content", null: false
+    t.date "date", null: false
+  end
+
+  create_table "proteome_caches", primary_key: "proteome_id", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
+    t.text "json_sequences", limit: 16777215, null: false
+  end
+
+  create_table "proteome_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
+    t.integer "uniprot_entry_id", null: false, unsigned: true
+    t.integer "proteome_id", limit: 3, null: false, unsigned: true
+  end
+
+  create_table "proteomes", id: :integer, limit: 3, unsigned: true, default: nil, force: :cascade do |t|
+    t.string "proteome_accession_number", limit: 12, null: false
+    t.string "proteome_name", limit: 145, null: false
+    t.integer "taxon_id", limit: 3, unsigned: true
+    t.binary "type_strain", limit: 1, default: 0b0, null: false
+    t.binary "reference_proteome", limit: 1, default: 0b0, null: false
+    t.string "strain", limit: 120
+    t.string "assembly", limit: 45
+    t.string "name", limit: 225
+  end
+
+  create_table "refseq_cross_references", id: :integer, unsigned: true, force: :cascade do |t|
+    t.integer "uniprot_entry_id", null: false, unsigned: true
+    t.string "protein_id", limit: 25
+    t.string "sequence_id", limit: 25
+  end
+
+  create_table "sequences", id: :bigint, unsigned: true, force: :cascade do |t|
     t.string "sequence", limit: 50, null: false
-    t.integer "lca"
-    t.integer "lca_il"
-    t.binary "fa"
-    t.binary "fa_il"
+    t.integer "lca", unsigned: true
+    t.integer "lca_il", unsigned: true
+    t.binary "fa", limit: 16777215
+    t.binary "fa_il", limit: 16777215
     t.index ["lca"], name: "idx_sequences_lca"
     t.index ["lca_il"], name: "idx_sequences_lca_il"
     t.index ["sequence"], name: "idx_sequences_sequence"
@@ -124,14 +156,14 @@
   create_table "taxons", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "name", limit: 120, null: false
     t.enum "rank", enum_type: "rank_type"
-    t.integer "parent_id"
+    t.integer "parent_id", limit: 3, unsigned: true
     t.integer "valid_taxon", limit: 2, default: 1, null: false
   end
 
   create_table "uniprot_entries", id: :integer, unsigned: true, force: :cascade do |t|
     t.string "uniprot_accession_number", limit: 10, null: false
-    t.integer "version", null: false
-    t.integer "taxon_id", null: false
+    t.integer "version", limit: 2, null: false, unsigned: true
+    t.integer "taxon_id", limit: 3, null: false, unsigned: true
     t.enum "type", null: false, enum_type: "db_type"
     t.string "name", limit: 150, null: false
     t.text "protein", null: false