1. DNA Identification: Stochastic Effects
Mark W Perlin, PhD, MD, PhD
Cybergenetics, Pittsburgh, PA
TrueAllele® Lectures
Fall, 2010
Cybergenetics ©
Cybergenetics ©

2. PCR is a random process COPY NO COPY
PCR efficiency is not 100% efficient. A strand copies with probability p, and doesn't copy with probability 1-p.
Cybergenetics ©

3. STR peak is a random variable
p = 80%, n = 6
2,900
1,298
One amplification
Another amplification
Cybergenetics ©

4. STR peak height measurement reflects probability distribution
mean = 2,145
stdev =
Cybergenetics ©

5. Relative peak certainty: coefficient of variation
stdev = 12
mean = 85
CV = 14%
standard
deviation
CV =
mean
value
Cybergenetics ©

6. Four times the peak height, gives twice the peak certainty
stdev = 25
mean = 350
CV = 7%
standard
deviation
CV =
mean
value
Cybergenetics ©

7. Peaks are probabilities
Cybergenetics ©

8. STR data is a random variable
Cybergenetics ©

9. Genotype pattern vs. peak data
Cybergenetics ©

10. Calculate stochastic effects
Computers can solve for genotype probabilities
and other random variables, like peak variation
By modeling peak variation
as just another parameter,
computers can calculate
DNA stochastic effects
Cybergenetics ©

11. Some apply a threshold Over threshold, peaks are treated
as allele events.
CPI model of data fit - all given equal likelihood. Previous threshold guidelines
Under threshold,
alleles do not exist.
list of
included alleles
Cybergenetics © 11

12. Since peaks are probabilities, thresholds introduce error
Cybergenetics ©

13. False allele exclusion rate

14. Probability preserves information
Cybergenetics ©

15. Data probability • arises from PCR randomness
• models stochastic effects
• helps explain allele drop out
• compares with genotype patterns
• preserves identification information
• established normative science
Cybergenetics ©