Arguments
fam.log.liks
the log likelihoods and lod scores for each family at each marker
(including the null hypothesis).
fam.log.liks is a 3-dimensional
matrix. The first dimension is indexed by the family identifiers.
The second dimension is indexed by the words
"log.lik" and
"lod.score". The third dimension is
indexed by the word "null" and the
names of the marker file names. To calculate the family log
likelihoods, calc.fam.log.liks = TRUE
must be passed
to multic via the
... parameter or a
multic.control object. If
fam.log.liks are not calculated, then
fam.log.liks is a character
vector providing instructions how to calculate the values.
fixed.effects
the estimate, standard error, t value, and p value of the fixed
effects for the traits and covariates for the null hypothesis and each
marker. fixed.effects is a 3-dimensional
matrix. The first
dimension is indexed by the trait and covariate names. The second
dimension is indexed by the words
"Estmate",
"Std.err",
"t.value", and
"p.value". The third dimension is
indexed by the word "null" and the
marker file names.
polygenic
the estimate, standard error, Wald score, Wald score P-value,
heritabilty estimate, standard error of the heritabilty
estimate, and heritably estimate P-value for the variance and
covariance of the polygenic effect of the formula
for the null hypothesis and each marker.
polygenic is
a 3-dimensional matrix. The first dimension is indexed by the letter
"s" followed by a
1, 2,
etc. for the variance of the first trait,
second trait, and so on or 12,
13, 23,
etc. for the covariance between the
first and second traits, first and third traits, second and third
traits, and so on. The second dimension is indexed by the words
"Estimate",
"Std.err",
"Wald",
"W.p.value",
"h^2",
"se.h^2", and
"h.p.value". The third dimension is
indexed by the word "null" and
the marker file names.
major.gene1
the estimate, standard error, Wald score, Wald score P-value,
heritabilty estimate, standard error of the heritabilty
estimate, and heritably estimate P-value for the variance and
covariance of the major gene effect of formula
for the null hypothesis and each marker.
major.gene1 is
a 3-dimensional matrix. The first dimension is indexed by the letters
"mg" followed by a
1, 2,
etc. for the variance of the first trait,
second trait, and so on or 12,
13, 23,
etc. for the covariance between the
first and second traits, first and third traits, second and third
traits, and so on. The second dimension is indexed by the words
"Estimate",
"Std.err",
"Wald",
"W.p.value",
"h^2",
"se.h^2", and
"h.p.value". The third dimension is
indexed by the word "null" and
the marker file names.
environmental
the estimate, standard error, Wald score, and Wald score P-value
for the variance and covariance of the environmental effect of formula
for the null hypothesis and each marker. environmental is
a 3-dimensional matrix. The first dimension is indexed by the letter
"e" followed by a
1, 2,
etc. for the variance of the first trait,
second trait, and so on or 12,
13, 23,
etc. for the covariance between the
first and second traits, first and third traits, second and third
traits, and so on. The second dimension is indexed by the words
"Estimate",
"Std.err",
"Wald", and
"W.p.value". The third dimension is
indexed by the word "null" and the
marker file names.
sibling.sibling
the estimate, standard error, Wald score, and Wald score P-value
for the variance and covariance of the sibling to sibling effect of formula
for the null hypothesis and each marker.
sibling.sibling is
a 3-dimensional matrix. The first dimension is indexed by the letters
"sib" followed by a
1, 2,
etc. for the variance of the first trait,
second trait, and so on or 12,
13, 23,
etc. for the covariance between the
first and second traits, first and third traits, second and third
traits, and so on. The second dimension is indexed by the words
"Estimate",
"Std.err",
"Wald", and
"W.p.value". The third dimension is
indexed by the word "null" and the
marker file names. To receive
valuable data, the 5th member of
constraints in the
multic.control
object must be set to not "F" (fixed).
parent.parent
the estimate, standard error, Wald score, and Wald score P-value
for the variance and covariance of the parent to parent effect of formula
for the null hypothesis and each marker.
parent.parent is
a 3-dimensional matrix. The first dimension is indexed by the letter
"p" followed by a
1, 2,
etc. for the variance of the first trait,
second trait, and so on or 12,
13, 23,
etc. for the covariance between the
first and second traits, first and third traits, second and third
traits, and so on. The second dimension is indexed by the words
"Estimate",
"Std.err",
"Wald", and
"W.p.value". The third dimension is
indexed by the word "null" and the
marker file names. To receive
valuable data, the 6th member of
constraints in the
multic.control
object must be set to not "F" (fixed).
parent.offspring
the estimate, standard error, Wald score, and Wald score P-value
for the variance and covariance of the parent to offspring effect of formula
for the null hypothesis and each marker.
parent.offspring is
a 3-dimensional matrix. The first dimension is indexed by the letter
"q" followed by a
1, 2,
etc. for the variance of the first trait,
second trait, and so on or 12,
13, 23,
etc. for the covariance between the
first and second traits, first and third traits, second and third
traits, and so on. The second dimension is indexed by the words
"Estimate",
"Std.err",
"Wald", and
"W.p.value". The third dimension is
indexed by the word "null" and the
marker file names. To receive
valuable data, the 7th member of
constraints in the
multic.control
object must be set to not "F" (fixed).
log.liks
the log likelihood, centimorgan distance, log likelihood status, and
lod score and P-value for the null hypothesis and each marker.
log.liks is a
data.frame. The row names are
"null" and the markder
file names. The column names are
"log.likelihood",
"distance",
"log.lik.status",
"lod.score", and
"p.value". The log likelihood
status represents whether the log likelihood converged before the
maximum interations allowed or not and have the values of either
"converg" or
"non-converg".
var.fixed
the variance of the fixed effects of the traits and covariates for the
null hypothesis and each marker.
var.fixed is a 3-dimensional
matrix. The first and second dimensions are indexed by the trait and
covariate names. The third dimension is indexed by the word
"null" and the
marker file names.
var.random
the variance of the polygenic, major gene, and
environmental effects for the null hypothesis and each marker.
var.random is a 3-dimensional matrix.
The first and second dimensions
are indexed as described by the polygenic, major.gene1, and
environmental components above. The third dimension is indexed by the
word "null" and the marker file names.
var.sandwich
a more precise variance estimator after using a sandwich estimator
approach. This is only calculated if the multic object represents a
univariate model. var.sandwich is a
3-dimensional matrix. The first
and second dimensions are indexed by
"s1",
"mg1", and
"e1". The third
dimension is indexed by the word "null"
and the marker file names.
cors
the Pearson, Spearman, genetic, environmental, and phenotypic
correlations. cors is a list made up of
the components "pearson",
"spearman",
"genetic",
"environment", and
"phenotype". Both
"pearson"
and "spearman" are their respective
correlations between the traits and
covariates. They are 2-dimensional matrices indexed by the trait and
covariate names. "genetic",
"environment", and
"phenotype" are the
respective correlations between the
polygenic and
environmenal
estimates. They are 2 dimensional matrices. The first dimension is
indexed by the word "null" and the
marker file names. The second
dimension is indexed as described by the covariance portions of the
polygenic and
environmenal components above.
v.matrices
the variance-covariance matrix of the trait (y) that incorporates the
polygenic, major gene, shared common environment, and error matrices.
v.matrices is a 2-dimensional matrix.
The first dimension is indexed
by the family identifier (famid) values.
The second dimension is
indexed by the word "null" and the
marker file names. Currently,
there are no individual identifiers on each of the V matrices. If the
V matrices are not calculated, then
v.matrices is a character vector
providing instructions how to calculate the values.
residuals
the observed values minus the fitted values of the trait (y) divided by
the square root of the V matrix for each family. If the residuals are
not calculated, then residuals is a
character vector providing
instructions how to calculate the values.
descriptives
the total individuals used, mean, standard deviation, minimum,
maximum, kurtosis, and skewness for each trait and covariate.
counts
various counts of the total number of pedigrees, people, females,
males, and so on. This is mostly for passing data for
print and
summary to display and is very likely to
be not useful to the user community.
call
how multic was called. call is a call
object.
R.sq
the proportion of variance due to the covariates.
metadata
a list of useful data like start.time,
finish.time,
call,
epsilon,
trait.count,
iterations,
null.initial.values,
method, etc.