PAGE 3


QUASI Principle Continues......

Mutiple aligments (625 sequences)

1   2  3   4   5   6  7  8   9  10  11
--- --- --- --- ttt --- --- --- --- --- ---
--- --- --- --- ttc --- --- --- --- --- ---
--- --- --- --- tat --- --- --- --- --- ---
--- --- --- --- ttt  --- --- --- --- --- ---
--- --- --- --- ttt  --- --- --- --- --- ---
--- --- --- --- ttt  --- --- --- --- --- ---
--- --- --- --- ttt  --- --- --- --- --- ---
--- --- --- --- ttt  --- --- --- --- --- ---
...................................................


(1) Buiding consensus

--- --- --- --- ttt  --- --- --- --- --- ---        Consensus

(2) Calculating  the observed  and expected  mutations

Amino Acide            Mutations          Obs      Exp R:S      Obs R:S

         Selection    
Phe (silent)                 ttt ---> ttc                10             na                na

Tyr                              ttt ---> tat (Tyr)       24           1:1             24:10

IIe                               ttt ---> att                  1           1:1               1:10

Leu                             ttt ---> tta                  3           3:1               3:10
                                              ttg
                                              ctt

Val                              ttt ---> gtt                  4           1:1               4:10

Ser                             ttt ---> tct                 23           1:1              23:10

Cys                            ttt ---> tgt                   0           1:1                0:10
               na

       Positive (p = 0.02)

       Negative (p = 0.01)

       Negative(p = 0.001)



       Neutral (p = 0.18)

       Positive (p = 0.04)

       Negative (p = 0.002)
                                                                                                                              derived from BMC Bioinformatics (2001) 2:1

(3) "The observed R:S is compared to neutral R:S by means of two-side test of bionomial distribution"(1)

       For Tyr at condon 5,

      (a) Calculate Sum of chances of r replacement mutations

            r = 24,                           (number of the observed replacement mutations)
            n = 24+10=34,             (number of mutations at this codon)
           p = 1/(1+1)= 0.5          (possiblity of Tyr muation at this condon)

         bTyr (24, 34, 0.5)=(3420)p24(1-p)(34-24)=0.0076    [ b(r,n,p)=(nr) pr (1-p)(n-r) ] ----chance of observed r replacement mutations

           Sum = bTyr (24, 34, 0.5) + bTyr (25, 34, 0.5) + bTyr (26, 34, 0.5) + ......... + bTyr (33, 34, 0.5)
                   = 0.0076 + 0.0031 + 0.0011 + ......... + 0.0000
                   = 0.01

           p = Sum x 2 = 0.02

     (b) Null hypothesis:  All replacement mutations are equally probable!  If p <0.05, reject null hypothesis

           since p < 0.05, Tyr is positively selected!

     (c) Repeat the calculations for IIe, Leu, Val, Ser and Cys

          ----- each p will be produced, if p < 0.05, this aa is positively selected!

          ------ the most frequent variants will be positively selected:
          ------ among the first round: 2H+ð
                             21
                 H = - € pi log2 pi             H: Shannon information
                       i=1                                            ð: standare error

           ------ Tyr and Ser



< previous
 1
 2
 3
 4
 5
 6
 7
 8
 9
 > next


Return to My Presentation
Return to My_Web_Site