Gene Expression analysis in complex diseases

1
Gene Expression analysis in complex diseases

• CHINDO HICKS, Ph.D.
• Department of Preventive Medicine and
  Epidemiology
• Bioinformatics Lecture for Graduate Students
• August 15th, 2007

2
Outline of lecture

• Part I
• microarray platforms and data processing
• Analysis to identify differentially expressed
  genes distinguishing disease from controls
• Example 1 Obese vs Lean subjects
• Example 2 Schizophrenia vs normal controls
• Analysis to identify genes distinguishing
  subtypes of cancer
• What is clustering?
• Example 3 classification of subtypes of cancer
• Example 4 Drug effectiveness

3
Outline of lecture

• Part II
• Bioinformatics tools for research
• Bioinformatics resources (databases)
• Problems for solving??
• Exam questions???

4
Part I Spotted microarray platform
5
Identification of differentially expressed genes
Disease
Control
Probe array
6
Gene expression data N54,000
7
Exampe 1 Identify genes that distinguish obese
from lean subjects in the Pima indian population
Lean 19
Obese 20
Fold change (FC) ?1 / ?2
T-test, null hypothesis H0 ?1 ?2
Alternative
hypothesis H1 ?1 ? ?2
Gene expression profiling based on RNA extracted
from the adipose tissue
8
RESULT Set of up and down regulated genes
distinguishing obese (OB) from lean (L) subjects
9
Example 2 Identify genes that distinguish
schizophrenia from control subjects
14 Matched controls
14 Schizophrenia
Fold change (FC) ?1 / ?2
T-test, null hypothesis H0 ?1 ?2
Alternative
hypothesis H1 ?1 ? ?2
Gene expression profiling based on RNA extracted
from the brain tissue
10
Result Set of genes distinguishing schizophrenia
(SCH) from controls (CNTL)
11
CLUSTERING What is a cluster?Identify genes
distinguishing the subgroups
12
Aim of clustering Group genes according to their
similarity

• Given genes x (x1, , xn), y (y1, , yn)

• Correlation distance

13
Hierarchical clustering

• Similarity of objects is represented in a tree
  structure (dendrogram).
• Advantage no need to specify the number of
  clusters in advance.
• Nested clusters can be represented.

14
Example 3. Classification of subtypes of ovarian
cancer
104 patients
Goal Find gene clusters distinguishing subtypes
of ovarian cancer
15
Clusters of genes distinguishing subtypes of
ovarian cancer
16
Major drugs ineffective for many
17
Cluster analysis of 118 anticancer drugs against
highly correlated genes
18
Part II Bioinformatics tools and resources for
research

• Bioinformatics tools
• Nucleotide Sequence Analysis
• Protein Sequence Analysis and Proteomics
• Structures
• Genome Analysis
• Gene Expression
• Tools for Programmers
• Resources
• We labs
• Databases

19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
Source http//www.ncbi.nlm.nih.gov/Tools/
25
http//homepage.univie.ac.at/herbert.mayer/AtoZ.ht
ml1to9
26
Problem solving

• (1) Use entrez (PubMed) or SWISS-PROT to find a
  protein sequence for the CD44 gene, download the
  sequence in FASTA format, use the sequence or
  gene name to and ProDOM database to find the
  domains for this gene
• (2) Use entrez (PubMed) or SWISS-PROT to find a
  protein sequence for the MMP-19 gene, down load
  sequence in FASTA format, use the sequence or
  gene name and Pfam database to find other members
  of the MMP family of proteins and domains
• Explain the functions of those domains