Arraylnformatics~(~R): An Integrated Microarray Lab Approach for Effective Data Analysis and Visualization

摘要

1. Quantity of data has outrun the capacity of first-generation analysis tools (spreadsheets): #1 microarray user complaint 2. Focused LIMS/Database solutions designed with intimate knowledge of the application process have become a requirement for multi-disciplined research efforts 3. General-purpose LIMS systems require a lot of customization ($) and often are not appropriate for smaller microarray research groups 4. Most microarray instruments are not designed to transfer directly with databases 5. Modest consensus emerging on best practices for analysis 5.1 Filter out low (noisy) signals, LOWESS normalize, Log-2 standardize across arrays, 2:1 ratio filtering, hierarchical clustering 5.2 But many researchers want to try different methods iteratively 5.3 So, flexibility to perform and examine multiple analyses is important 6. Experiment design is now raised much higher in priority and visibility: great analysis of bad data is not valuable.