In article <2547.9109101336 at crc.ac.uk> pwoollar at crc.ac.uk (Peter Woollard x3296) writes:
>With Lathrop & Lalouel's linkage analysis package (version 5.1),
>we already considering increasing the number of alleles from the
>defaults [currently 5 to 12 alleles depending upon the program]
>up to 20.
>>Have the authors of the above package ever considered converting
>over to C for the next release and also using dynamic memory
>allocation? (The latter would save having to recompile the
>programs, everytime a problem with different requirements is run.)
>
I agree this is a problem. You may be able to ameliorate it a bit by
going through the pedigree and manually reducing the number of alleles
by renaming alleles which are not involved in the same meiosis to have
the same number. If all individuals are typed then it should be possible
to only use four alleles for each marker, but this does not always work
if some parents are of unknown type. It would be helpful if there were
available a way to do this automatically. I don't know if such an
algorithm has been described??
As for using C - I have translated linkage 5.1 into C using the easily
available p2c translator without _too_ much difficulty. I think this has
enormous advantages in terms of increased portability, etc. I have been
running this version both on a Sun SPARCserver and also on a 486 under
DOS in 32-bit mode using the Zortech compiler with DOS extender. The
latter can also use virtual memory under Windows 3. Memory thus no
longer becomes a problem (anything too big is also too slow). I am
reluctant to release this translation for general use without the
express permission of the authors of LINKAGE, and anyway I do not feel
that I could competently support it. However I do think it would be both
feasible and extremely helpful if the people who currently support
LINKAGE could also provide a C translation of each program - as I say
this is not very difficult to produce at all once p2c is running and
configured OK. In this way development work could still be carried out
in Pascal but those users who wanted a C version could also obtain one.
Dave Curtis
Academic Department of Psychiatry, Janet: dcurtis at UK.AC.CRC
Middlesex Hospital, Elsewhere: dcurtis at CRC.AC.UK
Mortimer Street, London W1N 8AA. EARN/Bitnet: dcurtis%CRC at UKACRL
Tel 071-636 8333 Fax 071-323 1459 Usenet: ...!mcsun!ukc!mrccrc!D.Curtis
