Direct calibration of microarray
probes
Making a “DNA-meter”
Alex Pozhitkov, Peter A. Noble, Diethard Tautz,
Ploen Workshop, 2011.
Conundrum
• Signal intensity vs. GC
• Signal intensity vs. DG0
• Sense vs. antisense (same genomic loci) R2=0.37
Confirms:
Pozhitkov, et al (2006). Tests of rRNA hybridization … cannot be predicted. NAR
34, e66-e6
Melting study
•
•
•
•
Overhangs
Buffer composition
Tm Sense vs. Tm antisense
Tm vs. Signal Intensity on the array
Tm Antisense (0C)
Melting Temperatures
Tm Sense (0C)
Microarray is an Instrument
Instruments require calibration
“Good”, “Bad” and Noise
Signal Intensity
800
5000
4000
600
3000
400
2000
200
1000
0
2
4
8
6
0
2
Concentration
Individual replicas
Averages
4
6
8
Freundlich isotherm!
y=axb
Langmuir isotherm is not the only possible isotherm:
Pozhitkov et al. (2010) Beyond Affymetrix arrays: … NAR 38, e28.
DNA and RNA
y=axb+c
Measurements and Errors
Single probe replicate
Average over 10 replicates
CNV detection
CNV expected: 1
Sample prep variability
$

loci
Acknowledgements
• Max Planck Society
• Alabama State University

View