restrict. frag. freq. distrib.

Zharkikh Andrey gsbs1022 at utsph.sph.uth.tmc.edu
Fri Jan 15 14:31:47 EST 1993

``` eric at tomcat.onr.navy.mil ((Eric Eisenstadt)) writes:

>Is there a general and straightforward way to calculate the frequency
>distribution of different length restriction fragments generated by cutting
>a genome of size N with a specified set of restriction fragments?  For
>example, how many different 3 kb fragments are expected to result from
>digesting a 4 x 10E6 bp genome with two different six-cutters?

Let p is the frequency of six-cut sites.

[p=4^(-6)=1/4096 for equal base content].

Then the expected # of fragments of all sizes is M=N*p,
where L is the genome size.

[N=4,000,000;  M=977].

The probability of a fragment of size k > K is Prob(k>K)=(1-p)^k.

for K=3000, Prob(k>3000)=(1-1/4096)^3000=0.48

The expected number of such fragments is m=M*Prob(k>K)

m=977*0.48=469

For an interval of sizes: Prob(K1<k<K2)=Prob(k>K1)-Prob(k>K2).

say, for K1=3000 and K2=3100, Prob=0.48-0.47=0.01;

m=977*0.01=9.7

That's what you asked?

Andrey.

```