• Running Environment (for the four programs):    Sun OS version 5.7 (sun4u), 2624 MB RAM

• Considerable changes in running time should be expected on different computers. Runs on a modern PC are at least twice faster than the reported results.

• In the experiments we used Vitesse v1.0.  Comparison with Vitesse v2.0 will be added soon.

Experiment 1: (Same topology, increasing number of loci)
 

Experiment 2: (Same topology, increasing number of loci)
 

Experiment 3: (Different topologies)
 

Files	No. of People	No.of Loci	No. of Alleles	With Loops	Ln-Likelihood/ LodScore	Run  Time (Superlink)	Run Time (Fastlink)	Run Time (Vitesse)	Run Time (Genehunter)
datafileEC0/pedfileEC0	100	6	2	YES	-446.667268/0.014540	27.62	382.88	***	****
datafileEC1/pedfileEC1	100	7	2	NO	-506.520876/0.151373	2.43	0.82	0.40	****
datafileEC2/pedfileEC2	100	8	2	YES	-577.840308/-0.041029	0.56	12.59	***	****
datafileEC3/pedfileEC3	100	10	2	NO	-772.083821/-0.053660	16.36	**	22.79	****
datafileEC4/pedfileEC4	100	15	5-10	NO	-2918.451233/-0.186239	1.00	*	*	****
datafileEC5/pedfileEC5	20	15	2	NO	-248.830395/0.028368	44.1	*	*	****
datafileEC6/pedfileEC6	15	20	2	NO	-227.658948/-0.081496	35.32	*	*	****
datafileEC7/pedfileEC7	15	22	2	NO	-254.750791/-0.081557	102.56	*	*	****
datafileEC8/pedfileEC8	150	8	5-7	YES	-1896.360778/-0.55821	0.87	*	***	****
datafileEC9/pedfileEC9	150	10	5-7	YES	-2493.235175/-0.55821	1.28	*	***	****
datafileEC10/pedfileEC10	5	100	3-6	NO	-1007.835833/ 0.000067	0.06	**	*	0.41 (#)
datafileEC11/pedfileEC11	5	110	3-6	NO	-1103.826326/ 0.000067	0.08	**	*	0.45 (#)

Experiment 4: (Same topology, increasing number of loci)
 

            Clarifications to the tables:

*        Out-of-Memory
**      Segmentation Fault    /     Bus Error
***     Not Applicable - Vittese doesn't handle looped pedigrees.
****    Not Applicable - Genehunter doesn't handle large pedigrees.
*****   Over 100 hours.
#        In Genehunter, the lod-score is calculated several times, for  different positions of the disease gene on the marker map.
         In the runs above, the number of lod-score calculations equals the number of markers in the analsys (i.e, 99 for files
         datafileEC10.dat / pedfileEC10.dat  and 109 for files datafileEC11.dat / pedfileEC11.dat).
 
• Files   -    the test data files on which the programs were tested. The test data file for Experiment A is based on a pedigree taken from a coronary heart disease study.  The other test data files were generated by MakePed (a program I wrote).  MakePed generates random pedigree and loci data according to a specification given by the user.  The  user is required to enter the number of loci desired, the minimum and maximum number of alleles in each locus, the type of each locus being analyzed, the number of people in the pedigree and if loops and multiple marriages are allowed.  There is an option to generate a series of input files that are based on the same pedigree topology with increasing number of loci.
• No. of People  -   the number of people in the pedigree input file.
• No. of Loci  -   the number of loci being analyzed in the test data files.
• No. of Alleles  -   the number of alleles in each of the loci being analyzed.
• Includes Loops  -   indicates if there are inbreeding loops in the input pedigree.
• Run Time   -   running time (in seconds) of my program on the environment specified above.
• Run Time (Fastlink)  -  running time (in seconds) of Fastlink (mlink program) on the environment specified above.
• Run Time (Vitesse)   -  running time (in seconds) of Vitesse on the environment specified above.
• Ln-Likelihood  -  the ln-likelihood of the given input pedigree, given the recombination values between each two consecutive loci.