IUBio

LINKAGE bug

Jurg Ott OTTSR%CUVMB at CUVMB.CC.COLUMBIA.EDU
Tue Sep 14 07:39:28 EST 1993


Jurg Ott, Columbia University, New York
jurg.ott at columbia.edu
14 September 1993


          Report of problems in LINKAGE version 5.20

     Two researchers (Drs. Ken Morgan and Marcella Devoto) have
alerted me to problems in the newest version of the LINKAGE
programs.  The example given below shows an analysis done with
MLINK on five families and, in the far right column, analogous
results when families 2 and 3 are run separately each.  The arrow
pinpoints the problem areas;  "-inf" indicates a zero likelihood.

     No problem occurs when the data are analyzed with LINKAGE
version 5.10.  I have made Mark Lathrop aware of this situation
and he is in the process of fixing the bugs.  Until these prob-
lems are resolved we are sending out version 5.10 of the LINKAGE
programs.

Output:

Length of real variables = 8 bytes
LINKAGE (V5.20) WITH  2-POINT AUTOSOMAL DATA
 ORDER OF LOCI:   1  2

THETAS  0.500                         LOG 10 LIKE with each
PEDIGREE |  LN LIKE  | LOG 10 LIKE    fam. run separately
-----------------------------------   -------------------
        1   -72.260907   -31.382446
        2   -50.320901   -21.854043   -21.854043
        3   -31.553232   -13.703365   -13.703365
        4   -76.959775   -33.423134
        5   -22.809778    -9.906140
-----------------------------------

THETAS  0.000
PEDIGREE |  LN LIKE  | LOG 10 LIKE
-----------------------------------
        1   -71.736675   -31.154776
        2   -50.035346   -21.730028   -21.730028
        3   -34.997827   -15.199331   -15.199331
        4   -78.074510   -33.907257
        5   -23.013923    -9.994799
-----------------------------------

THETAS  0.010
PEDIGREE |  LN LIKE  | LOG 10 LIKE
-----------------------------------
        1   -71.745146   -31.158454
        2   -50.024837   -21.725464   -21.725464
        3   -34.581294   -15.018433   -15.018433
        4   -78.018161   -33.882784
        5   -23.004994    -9.990921
-----------------------------------

THETAS  0.050
PEDIGREE |  LN LIKE  | LOG 10 LIKE
-----------------------------------
        1   -71.784345   -31.175479
        2   -52.555550   -22.824537   -22.856760  <==
        3         -inf         -inf   -14.634118  <==
        4         -inf         -inf
        5         -inf         -inf
-----------------------------------

THETAS  0.100
PEDIGREE |  LN LIKE  | LOG 10 LIKE
-----------------------------------
        1   -71.841754   -31.200411
        2   -52.510950   -22.805167   -22.834461  <==
        3         -inf         -inf   -14.382773  <==
        4         -inf         -inf
        5         -inf         -inf
-----------------------------------

THETAS  0.200
PEDIGREE |  LN LIKE  | LOG 10 LIKE
-----------------------------------
        1   -71.969734   -31.255992
        2   -52.541714   -22.818528   -22.844912  <==
        3         -inf         -inf   -14.092111  <==
        4         -inf         -inf
        5         -inf         -inf
-----------------------------------



More information about the Gen-link mailing list

Send comments to us at biosci-help [At] net.bio.net