Reverse Transcription

This step simulates the reverse transcription (RT) and an optional fragmentation or filtering process which can influence the outcome of the run substantially. You may either choose a random priming strategy or poly-dT primers (RT_PRIMER), but you have to provide the maximum and minimum length of the expected RT products (RT_MIN, RT_MAX) — which certainly are a function of the specific RT protocol (enzymes and timing) applied. If the RT protocol is POLY-DT primed, a single successful priming event on each RNA molecule is assumed — which probably does not reflect reality, but prevents from a statistical even spread of molecule loss. In the case of RANDOM priming, the number of successful priming events, i.e., primers that recruit a reverse transcriptase, for a certain $RNA_x$ molecule is drawn from sampling poisson distribution with mean

(1)
\begin{align} \mu= \frac{n* length(RNA_x)}{\sum_{i=1}^n length(RNA_i)} \end{align}

Eq.(1) compares $length()$ of the considered molecule $RNA_x$ to the average length of the RNA molecules in the reaction. To prevent the loss, especially of shorter transcripts, that may incur due to molecule numbers that are lower than in real experiments, at least one successful priming event is enforced on every RNA molecule. The length of the generated RNA molecule then is determined by an uniform randomly distributed variable $U= [0;1[$.

(2)
$$RT_{min} + U * (RT_{max}- RT_{min})$$

In the case of multiple priming events are extended along the same RNA molecule, upstream primed first strand cDNA synthesis can displace further downstream bound primers. The chance of for a DNA-RNA hybrid to be displaced is most likely a function of its distance $dist$ to the closest upstream priming event. The FLUX SIMULATOR currently decides on the displacement of overlapping RT extensions in a Bernoulli trial. To be specific, downstream primers get displaced if for a uniformly drawn random variable $U=[0;1[$ holds

(3)
\begin{align} U> \frac{dist}{RT_{min}}- 1 \end{align}