Relational Databases: Basic Concepts

BCHB5242015

Lecture 21

By Edwards & Li

Slides: https://goo.gl/rl1wFL

Outline

What is a (relational) database? When are relational databases used? Commonly used database management

systems Using existing databases Creating and populating new databases Python and relational databases Exercises

(Relational) Databases

Databases store information Bioinformatics has lots of file-based information:

FASTA sequence databases Genbank format sequences Store sequence, annotation, references, annotation Good as archive or comprehensive reference Poor for a few items

Relational databases also store information Good for a few items at a time Flexible on which items

Relational Databases

Store information in a table Rows represent items Columns represent items' properties or attributes

Name Continent RegionSurface

AreaPopulatio

nGNP

BrazilSouth

America South America 8547403 170115000 776739

Indonesia Asia Southeast Asia 1904569 212107000 84982

India AsiaSouthern and Central

Asia 3287263101366200

0 447114

China Asia Eastern Asia 9572900127755800

0 982268

Pakistan AsiaSouthern and Central

Asia 796095 156483000 61289

United States

North America North America 9363520 278357000 8510700

Relational Databases

Tables can be millions of rows Can access a few rows fast

Countries more than 100,000,000 in population? Countries on the “Asia” continent? Countries that start with “U”? Countries with GNP = 776739

Name Continent RegionSurface

AreaPopulatio

nGNP

BrazilSouth

America South America 8547403 170115000 776739

Indonesia Asia Southeast Asia 1904569 212107000 84982

India AsiaSouthern and Central

Asia 3287263101366200

0 447114

China Asia Eastern Asia 9572900127755800

0 982268

Pakistan AsiaSouthern and Central

Asia 796095 156483000 61289

United States

North America North America 9363520 278357000 8510700

When are Relational Databases Used? LARGE datasets

Does data fit in memory? Store data first ...

... ask questions later Lookup or sort by many keys

For single key, simple data structures often work Store results of expensive compute or data-cleanup

Compute once and return results many times "Random" or unknown access patterns Specialized data-structures not appropriate

Use string/sequence indexes for sequence data

What is RDBMS?

• RDBMS stands for Relational Database Management System.

• RDBMS is the basis for SQL, and for all modern database

• The data in RDBMS is stored in database objects called tables.

• A table is a collection of related data entries and it consists of columns and rows.

Common DBMS

Oracle Commercial, market leader, widely used in

businesses MySQL

Free, open-source, widely used in bioinformatics, suitable for large scale deployment

Sqlite Free, open-source, minimal installation

requirements, no users, suitable for small scale deployment

http://db-engines.com/en/ranking

What is SQL?

• SQL stands for Structured Query Language

• SQL lets you access and manipulate databases

• SQL is an ANSI (American National Standards Institute) standard

• Although SQL is an ANSI (American National Standards Institute) standard, there are different versions of the SQL language.

http://www.w3schools.com/sql/sql_intro.asp

What Can SQL do?

• SQL can execute queries against a database

• SQL can retrieve data from a database

• SQL can insert records in a database

• SQL can update records in a database

• SQL can delete records from a database

• SQL can create new databases

• SQL can create new tables in a database

• SQL can create stored procedures in a database

• SQL can create views in a database

• SQL can set permissions on tables, procedures, and views

http://www.w3schools.com/sql/sql_intro.asp

Lets look at some examples

We'll use a third-party program to "look at" Sqlite databases: SqliteStudio (Linux), SqliteSpy (Windows), …

Download examples: World.db3, taxa.db3 from Course data folder

Use SqliteStudio to look at examples World.db3, taxa.db3

Using existing databases

Use the "select" SQL command to find relevant rows select * from Country where Population > 100000000; select * from Country where Continent = 'Asia'; select * from Country where Name like 'U%'; select * from Country where GNP = 776739;

Each command ends in semicolon ";". "where" specifies the condition/constraint/rule. "*" asks for all attributes from the relevant rows. Lets experiment with world and taxa databases.

Using existing databases

Select can combine (“join”) multiple tables Use the where condition to match rows from each

table and “link” corresponding rows…

select * from taxonomy, name where taxonomy.rank = 'species' and name.name_class = 'misspelling' and name.tax_id = taxonomy.tax_id

Using existing databases

Select can sort and/or return top 10

select * from taxonomy limit 10;

select * from taxonomyorder by scientific_name;

select * from taxonomyorder by tax_id desclimit 10;

Using existing databases

Select can count and do string matching.

"like" uses special symbols: % matches zero or more symbols _ match exactly one symbol

Some RDBMS support regular expressions MySQL, for example.

select count(*) from taxonomy where scientific_name like 'D%';

Creating databases

Use the "create" SQL command to create tablesCREATE TABLE taxonomy ( tax_id INTEGER PRIMARY KEY, scientific_name TEXT, rank TEXT, parent_id INT);CREATE TABLE name ( id INTEGER PRIMARY KEY, tax_id INT, name TEXT, name_class TEXT);

Populating databases

Use the "insert" SQL command to add rows to tables Usually, the special id column is initialized

automatically

INSERT INTO name (tax_id,name,name_class) VALUES (9606,'H. sapiens','synonym');

SELECT * from name where tax_id = 9606;

Python and Relational Databases

Issue select statements from python and iterate through the results

Sometimes it is easiest to make Python do some of the work!

import sqlite3conn = sqlite3.connect(‘taxa.db3')c = conn.cursor()c.execute("""   select * from name   where name like 'D%'   limit 10; """)for row in c:   print row

Python and Relational Databases

Use parameter substitution for run-time values

import sysimport sqlite3

tid = int(sys.argv[1])

conn = sqlite3.connect('taxa.db3')params = [tid,'scientific name']c = conn.cursor()c.execute("""   select * from name   where tax_id = ? and name_class = ?;""",params)for row in c:   print row

Next-time: Object-relational mappers

Setup python to treat tables as classes, rows as objects

# Set up data-modelfrom model import *

hs = Taxonomy.get(9606)for n in hs.names:    print n.name, "|", n.nameClass

condition = Name.q.name.startswith('Da')for n in Name.select(condition):    print n.name, "|", n.nameClass

Lab exercises

Read through an online course in SQL sqlcourse.com, sql-tutorial.net, ...

Write a python program to lookup the scientific name for a user-supplied organism name.