A stochastic model of the processes in PCR based amplification of STR DNA in forensic applications

Schematic presentation of the stochastic model. During a complete cycle, there are three different possibilities with an individual strand, each with a certain probability.

Mean number of chains as function of the cycle number. The numbers near the lines reflect the number of stutters. The solid lines reflect the calculated exact results, the dashed lines the approximated results, and the symbols the simulated results. On the current scale, the approximation is not visible for chains with 2 stutters.

The ratios of the expected numbers of chains with no, 1 or 2 stutters. The numbers near the lines indicate the ratio plotted.

Coefficients of variation (expressed as percentage) for the chains with no, 1 or 2 stutters (symbols 0, 1 and 2, respectively). No approximation is available for 2 stutters.

The correlation between the numbers of chains with no and 1 stutter (symbol 0,1, including approximation), with no and 2 stutters (symbol 0,2), and with 1 and 2 stutters (symbol 1,2).

The bivariate empirical probability distributions as obtained in the computer simulations after 28, 31 and 34 cycles. The lines correspond to the situations that the stutter peak intensity is 6%, 7.5% and 15% of the allelic peak intensity.