| [ < ] | [ > ] | [ << ] | [ Up ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Only functions affecting general MORGAN
capabilities or enabling new
programs are listed here. Many other general improvements in
Morgan library functions and MORGAN
set-up routines have been made.
Some details may be found in MORGAN
web release notes, or in the
README_relnotes of each MORGAN
release.
In MORGAN V2.8.2, forward HMM computation for multiple meioses has been replaced by a factored version (FHMM), enabling much faster exact computation on small pedigree components and multiple-meiosis sampling for larger numbers of meioses.
Exact computation of lodscores on small pedigree components
has been implemented for lm_markers: computation uses the FHMM
version of the Baum algorithm.
MM-sampler updates multiple meioses jointly and is therefore a generalization of the meiosis sampler (M-sampler). There are four types of update in MM-sampler: random meiosis update, individual update, sib update and 3-generation update. This is based on work by Liping Tong.
For more detail on L- and M-samplers, see Using MCMC to Estimate Parameters of Interest in Pedigree Data.
Up to MORGAN V2.8.2, MCMC was performed globally over pedigree components (except those small enough for exact computation). The L-sampler peeling and lod score estimation could be done either by component (using "set peeling by component") or globally (the default).
With MORGAN V2.8.3, and specifically to accommodate the new
lm_haplotype program, the preferred option is to do both MCMC and
pedigree peeling (lod score estimation) by component, and to use exact
computation on all sufficiently small component pedigrees. The
alternative, retained so that older data sets can be rerun, is to
use "set global MCMC", in which case no exact computation will be
done, and MCMC will be done globally over all component pedigrees.
In this case, the "set peeling by component" option is retained.
As yet, loci are either multiallelic marker loci assumed observed
without error, or trait loci which may have general penetrance functions
but are diallelic. In order to allow models for "non-genotypic"
markers, general joint peeling programs have been implemented, based
on Thompson (1976: UU Tech Rept, #6). These peeling routines are used
by the lm_map program which allows for errors in marker data. They are
not yet released, as they are still in process of testing: they may be
released in MORGAN V2.8.3.
In MORGAN V2.8.2,
liability penetrances (previously available only for lm_bayes)
have been implemented for lm_markers. Penetrances for each
liability class are now read from an input file using the
"input extra data file S" parameter statement.
Additionally, an age-based penetrance function for a qualititative trait has been implemented. That is, penetrances are directly dependent on age, rather than going through a liability class specification.
In two-locus models for a quantitative trait, penetrances may be specified as additive, with a genotypic mean for each trait genotype for each locus. Alternatively, a matrix array of 2-locus genotypic means may be specified, allowing for epistasis (see Sung & Wijsman, 2007, Human Heredity 63: 144-152.).
With more complex trait models, including those of
lm_twoqtl
(see Sung et al., 2007, Genetic Epidemiology 31: 103-114), a more general
specification of traits is required. In MORGAN V3.0,
completely new structures have been introduced, separating traits
(phenotypes) from trait loci ("tlocs"). Traits may be affected by
genotypes at several tlocs; the genotypes at a tloc may affect several
traits. This more general structure is not yet released -- see
lm_twoqtl below.
lm_map with and without error models:
Some updates and corrections to the lm_map program are made
in MORGAN V2.8.2. Additonally under development is a version
of lm_map which allows for error in observation of marker
genotypes. This version may be released with MORGAN V2.8.3.
The version of the program lm_pval released in
MORGAN V.2.8 and subsequent, and described in this tutorial,
uses the latent p-value distribution of Thompson & Geyer (2007,
Biometrika). Additional programs using these ideas are under
development, including programs for the distribution of latent lod
scores obtained in MCMC sampling (lm_fuzlod),
p-values and randomized tests
based on latent lod score statistics (lm_fzplod),
and randomized confidence sets
for the location of a trait locus (lm_fzconf). These are working
names only. The methods are described in Thompson (2007: submitted).
Gold replaces the previous
Gold2
subdirectory, for tests of
lm_auto, lm_pval,
lm_map and lm_ibdtests.
Gold1 lm_auto tests remain temporarily,
since they provides the only tests of MCMC samplers on looped
pedigrees. Gold1 lm_auto
gold standards were omitted from the released MORGAN
V2.8.2, due to delays in checking looped pedigree peeling routines:
they will be reinstated in MORGAN V2.8.3.
To make room for new Lodscore programs being released in
MORGAN V2.8.2, 2.8.3 and 3.0, the two older programs
lm_schnell and lm_lods have been moved to the
new directory LR_Lods. These two programs differ in several ways
from newer programs, but the principal one is that they use the
methods of combining likelihood ratios (LR) along the chromosome
in order to estimate lod scores (see Thompson & Guo, 1991, IMA J
Math Appl in Med & Biol).
The Gold1 subdirectory remains temporarily,
since it provides the only tests of MCMC samplers on looped
pedigrees. Gold1 lm_lods
gold standards were omitted from the released MORGAN
V2.8.2, due to delays in checking looped pedigree peeling routines:
they will be reinstated in MORGAN V2.8.3.
lm_multiple and lm_markers
A new lod score calculation program lm_multiple was released in
MORGAN V2.8.2 (Spring, 2006). The lm_markers program
is still made as a separate executable, but is compiled as a special
case of lm_multiple code. In V2.8.2,
both programs perform exact lodscore
computations on small pedigree components. In V2.8.3 (unreleased) this
is optional: see lm_haplotype below.
The difference between lm_multiple and lm_markers is that
the new lm_multiple substitutes the old single-meoisis M-sampler
updates for the new multiple-meiosis (MM) sampler that is
the work of Liping Tong (Tong & Thompson, 2007, Human Heredity:in
press): see above.
lm_multiple runs with the same parameter file and other input
files required
by lm_markers. The output is also essentially the same as that from
lm_markers.
More information on lod score calculation programs in MORGAN V.2.8.1 (and previous) can be found in Estimating Location LOD Scores by MCMC.
lm_haplotype
The lm_haplotype program is a generalization of
lm_multiple in which haplotypes of key individuals dividing the
pedigree are sampled in addition to meiosis indicators. To facilitate
efficient implementation of this algorithm, new peeling-by-component
routines need to be implemented and checked. This program is the work
of Liping Tong (Tong & Thompson, 2007, Human Heredity:in
press). The program is in process of release: probably in
MORGAN V2.8.3.
lm_multiple
are added in the Lodscore/Gold2 subdirectory.
In MORGAN
V2.8.3, the Gold directory
replaces the previous Gold2 directory. Gold directories for
lm_lods
and lm_schnell are moved to the new LR_Lods program
directory. Thus Gold1 no longer exists in Lodscore.
lm_twoqtl
The new program lm_twoqtl allows two (linked or unlinked)
quantitative trait loci to contribute additively or epistatistically
to a single trait (see Sung et al., 2007, Genetic Epidemiology
31: 103-114).
An unreleased version of this program exists under MORGAN
V2.8, but a user-friendly version requires a new MORGAN
structure in which traits are separated from trait loci or "tlocs": see
above. This new structure is implemented in
MORGAN V3.0. The program lm_twoqtl will be the first
to be released under this structure, although eventually all current
programs will be converted.
| [ < ] | [ > ] | [ << ] | [ Up ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |