Quality Control of Gene Expression Data
Introduction
High-throughput gene expression platforms are powerful techniques but are quite susceptible to variations, either be from biological or technical sources. This is a non-exhaustive list of possible variation sources:
Biological Variability
- Genetic variations
- Biomaterial sampling
- Quality/quantity of the extracted mRNA
Technical Variability
- mRNA processing (i.e.: cDNA synthesis)
- Fluorophore incorporation
- Array quality
- Hybridization
- Image acquisition
Because of this, it is best to verify before hand the quality of the data before getting into the normalization/summeraization process and the following analysis process. Replicates showing excessive variations will induce problems afterward, creating excessive noise in the data.
Mechanically-spotted arrays: quality control
GUI methods
- More to come: using SpotFinder
CLI methods
Affymetrix: Quality Control
GUI Methods
- More to come: using RMAExpress to get NUSE and RLE graphs
CLI methods
Illumina: Quality Control
- More to come…