Usage of VITESSE

Alex Schaffer schaffer at cs.rice.edu
Tue Nov 18 11:12:59 EST 1997


This is a follow-up to the Iain Fenton's and Ellen Wijsman's thread about
VITESSE.

I concur that when VITESSE is applicable (simple pedigrees, MLINK
or LINKMAP mode) it should be preferred to FASTLINK. FASTLINK
does also provide ILINK mode which allows for optimizion of
the recombination fraction and/or allele frequencies, neither
of which is currently available in VITESSE.
Pieces of FASTLINK are also used in the simulation program FASTSLINK
and the genotype checking program GenoCheck for which there are
no VITESSE counterparts,

Ellen Wijsman's statement that the number of alleles does not
matter to VITESSE is an oversimplification. It matters
much less to VITESSE than to FASTLINK and does not matter for
some combinations of pedigree structure and set of genotyped
individuals.

At NIH, we have successfully used a FASTLINK/VITESSE combination
to do LINKMAP on data sets that had a mixture of non-simple pedigrees and
simple pedigrees. We use VITESSE on the simple
pedigrees, FASTLINK/LINKMAP on the non-simple pedigrees and
then merge results using a perl script prepared by Jim Tomlin 
(jtomlin at helix.nih.gov).   

As for acceptability in print, my count suggests that over 10%
of 1997 linkage analysis papers are using VITESSE at least for some analyses
and some of these are in the best journals (e.g., the recent paper
in JAMA on Alzheimer's linkage to chr. 12). 

However, it is a sad, sad state of affairs that over 50% of
linkage analysis papers still use only LINKAGE.




More information about the Gen-link mailing list