Some statistics for DNA chip data

11/10/2002

Click here to start

Table of Contents

Some statistics for DNA chip data

Acknowledgments

Five statistical issues

There are many other general issues, more bioinformatic than statistical

Some terminology

Two-channel microarrays

High density short oligonucleotide arrays

One probe set (˜gene ) ?11-20 PM,MM pairs

Summarizing and removing artefacts from raw data

Quantification of expression: cDNA arrays

Some local backgrounds

15 replicates of a high responding Affymetrix probe set: PM blue, MM red

Quantification of expression: high densityoligonucleotide (Affymetrix) arrays

The need for robustness: 2 probe sets

The use of different scanners

Same cDNA slide, different scanners

All slides: by scanner

Same cDNA slide, different image analyses(includes background adjustment)

Spatial bias in cDNA arrays

Intensity-related spatial biases

Designing gene expression experiments

Natural design choice

Treatment versus Control

Types of replication

Technical replication - labeling

Biological replicates

Biological replicates: effect of sample size

Use of dye-swaps

Replication: a summary

Which genes are differentially expressed?

Some approaches in use

Est. age effects vs their est. SDs

A 2?2 factorial with Affymetrix chips

Which genes exhibit interesting expression patterns?

Can we discover novel pathways?

Author: WEHI ITS