Executables required: mfmap, mflink, unknown, noscore.
Although standard lod score analysis can be a powerful means of detecting linkage, it is necessary to specify a transmission model for the disease locus. If this transmission is specified wrongly and does not match the true mode of transmission then the lod scores produced may only be small, or even negative, even when the marker tested is in fact linked to the disease gene. A variety of methods have been developed which aim to avoid this problem. The approach implemented in mflink is to calculate the log likelihood for a range of transmission models. Unlike classical lod score analysis, in which the recombination fraction between the affection and marker loci is altered, mflink tests one specific position but alters the transmission model used. The model is defined by one parameter, the heterozygote penetrance, which is varied to produce a range of models from Mendelian dominant to no effect to Mendelian recessive. Log likelihoods are calculated assuming that the disease locus is at the test position or is at an unlinked position. Admixture is incorporated, and the proportion of pedigrees which may contain a mutation at the test position is also varied. The mflink program calculates three statistics which measure the evidence in favour of a disease gene being present at or near the test position:
Setting up mflink analyses is made considerably easier by using the mfmap utility which is provide in the mflink package. The mfmap program can use a standard pedigree and locus data file in LINKAGE format and run an mflink analysis between the affection locus and each marker locus in the files. The results of all the analyses are summarised in a single output file. The mfmap program requires that the pedigree and locus data files contain a single affection locus followed by all the required marker loci.
We will use the same files as we previously used for the model-specific analyses, called alzall.ppd and alzall.par. These are the files which contain a description of the Alzheimer's disease locus and three marker loci, but which do not contain the locus with age-specific liability classes. The only modifications which need to be made before these files can be used by mfmap is to specify the correct marker map.
Modify alzall.par with a text editor. For mfmap it is important that the marker loci are specified as being in the correct order, so change the line providing the locus order to read as follows:
1 2 4 3 << order of lociIt is also important to provide the correct recombination fractions between the marker loci, so change the penultimate line of the file, which provides the recombination fractions to read as follows:
0.5 0.04 0.06What these changes imply is that the markers lie in order MAR1-MAR3-MAR2 with a recombination fraction of 0.04 between MAR1 and MAR3 and a recombination fraction of 0.06 between MAR3 and MAR2.
When these changes have been made alzall.par should appear as follows:
4 0 0 5 << no loci, risk locus, sexlinked(if 1) 0 0.0 0.0 0 << mut locus, mut rate, haplotype freq(if 1) 1 2 4 3 << order of loci 1 2 # ALZ 0.9999 0.0001 << gene freqs 1 << number of liability classes 0.01 0.5 0.5 3 4 # MAR1 0.14 0.32 0.21 0.33 << gene freqs 3 3 # MAR2 0.4 0.4 0.2 3 3 # MAR3 0.3 0.4 0.3 0 0 0.5 0.04 0.06 1 0.05 0.4
Save the file.
Now that we have a pedigree and locus data file describing one affection locus and three marker loci, we can test the affection locus against each marker in turn using the mfmap utility. At the system prompt enter:
mfmap alzall.ppd alzall.par alzallmf.out alzallmf.log
This will carry out a two-point analysis with each marker, with the test position set at a recombination fraction of 0.05 between the disease and marker loci. When the program has finished running you can view the output, which will be written to alzallmf.out. Examine this file with a text editor. Note the values of the different lod scores produced and which marker seems to provide the strongest evidence for linkage. The log file, alzallmf.log, can also be examined with a text editor and contains the detailed results from each mflink analysis. It lets one see information such as which transmission model seems most compatible with linkage.
As well as carrying out two-point analyses with each marker, mfmap will also allow automatically carry out multipoint analyses to test each interval between markers. As an example of this facility, at the system prompt enter:
mfmap alzall.ppd alzall.par alzmf3p.out alzmf3p.log -n1
The -n1 (the digit 1 not the letter L) parameter defines the number of markers used on either side to flank the test interval. Here, we have one marker on either side along with the affection locus, meaning that a three-point analysis is performed. The test position is set to be between midway between each pair of markers. Additional positions are tested at either end of the marker map, at a recombination fraction of 0.05 with the end markers. Again, the files alzmf3p.out and alzmf3p.log can be examined with a text editor to see which interval provides most support for linkage to the disease.
This section demonstrates the application of one "model-free" approach to the analysis of diseases with a complex mode of inheritance, using mfmap and mflink to test a disease locus against different positions on a marker map.
Exercises in genetic linkage analysis