Package 'Pedixplorer'

Description

The Pedixplorer package for pedigree data an updated package of the kinship2 package. The kinship2 package was originally written by Terry Therneau and Jason Sinnwell. The Pedixplorer package is a fork of the kinship2 package with additional functionality and bug fixes.

Details

The package download, NEWS, and README are available on CRAN: \urlhttps://cran.r-project.org/package=kinship2 for the previous version of the package.

Functions

Below are listed some of the most widely used functions available in arsenal:

Pedigree(): Contstructor of the Pedigree class, given identifiers, sex, affection status(es), and special relationships

kinship(): Calculates the kinship matrix, the probability having an allele sampled from two individuals be the same via IBD.

plot() : Method to transform a Pedigree object into a graphical plot. Allows extra information to be included in the id under the plot symbol. This method use the plot_fromdf() function to transform the Pedigree object into a data frame of graphical elements, the same is done for the legend with the ped_to_legdf() function. When done, the data frames are plotted with the plot_fromdf() function.

shrink(): Shrink a Pedigree to a specific bit size, removing non-informative members first.

bit_size(): Approximate the output from SAS's ⁠PROC FREQ⁠ procedure when using the ⁠/list⁠ option of the TABLE statement.

Data

• sampleped(): Pedigree example data sets with two pedigrees

• minnbreast(): Larger cohort of pedigrees from MN breast cancer study

Author(s)

Maintainer: Louis Le Nézet [email protected] (ORCID)

Authors:

• Jason Sinnwell [email protected]

• Terry Therneau

Other contributors:

• Daniel Schaid [contributor]

• Elizabeth Atkinson [contributor]

• Louis Le Nezet [contributor]

See Also

Useful links:

• https://github.com/LouisLeNezet/Pedixplorer

• Report bugs at https://github.com/LouisLeLezet/Pedixplorer/issues

Examples

library(Pedixplorer)

Description

Given a Pedigree, this function creates helper matrices that describe the layout of a plot of the Pedigree.

Usage

## S4 method for signature 'Pedigree'
align(
  obj,
  packed = TRUE,
  width = 10,
  align = TRUE,
  hints = NULL,
  missid = "NA_character_"
)

Arguments

Details

This is an internal routine, used almost exclusively by ped_to_plotdf().

The subservient functions auto_hint(), alignped1(), alignped2(), alignped3(), and alignped4() contain the bulk of the computation.

If the hints are missing the auto_hint() routine is called to supply an initial guess.

If multiple families are present in the obj Pedigree, this routine is called once for each family, and the results are combined in the list returned.

For more information you can read the associated vignette: vignette("pedigree_alignment").

Value

A list with components

• n: A vector giving the number of subjects on each horizonal level of the plot

• nid: A matrix with one row for each level, giving the numeric id of each subject plotted. (A value of 17 means the 17th subject in the Pedigree).

• pos: A matrix giving the horizontal position of each plot point

• fam: A matrix giving the family id of each plot point. A value of 3 would mean that the two subjects in positions 3 and 4, in the row above, are this subject's parents.

• spouse: A matrix with values

  • 0 = not a spouse

  • 1 = subject plotted to the immediate right is a spouse

  • 2 = subject plotted to the immediate right is an inbred spouse

• twins: Optional matrix which will only be present if the Pedigree contains twins :

  • 0 = not a twin

  • 1 = sibling to the right is a monozygotic twin

  • 2 = sibling to the right is a dizygotic twin

  • 3 = sibling to the right is a twin of unknown zygosity

See Also

alignped1(), alignped2(), alignped3(), alignped4(), auto_hint()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
align(ped)

Description

First alignment routine which create the subtree founded on a single subject as though it were the only tree.

Usage

alignped1(idx, dadx, momx, level, horder, packed, spouselist)

Arguments

Details

In this routine the nid array consists of the final ⁠nid array + 1/2⁠ of the final spouse array. Note that the spouselist matrix will only contain spouse pairs that are not yet processed. The logic for anchoring is slightly tricky.

1. Anchoring:

First, if col 4 of the spouselist matrix is 0, we anchor at the first opportunity. Also note that if spouselist[, 3] == spouselist[, 4] it is the husband who is the anchor (just write out the possibilities).

2. Return values initialization:

Create the set of 3 return structures, which will be matrices with 1 + nspouse columns. If there are children then other routines will widen the result.

3. Create lspouse and rspouse:

This two complimentary lists denote the spouses plotted on the left and on the right. For someone with lots of spouses we try to split them evenly. If the number of spouses is odd, then men should have more on the right than on the left, women more on the right. Any hints in the spouselist matrix override. We put the undecided marriages closest to idx, then add predetermined ones to the left and right. The majority of marriages will be undetermined singletons, for which nleft will be 1 for female (put my husband to the left) and 0 for male. In one bug found by plotting canine data, lspouse could initially be empty but length(rspouse) > 1. This caused nleft > length(indx). A fix was to not let indx to be indexed beyond its length, fix by JPS 5/2013.

4. List the children:

For each spouse get the list of children. If there are any we call alignped2() to generate their tree and then mark the connection to their parent. If multiple marriages have children we need to join the trees.

5. Splice the tree:

To finish up we need to splice together the tree made up from all the kids, which only has data from lev + 1 down, with the data here. There are 3 cases:

1. No children were found.

2. The tree below is wider than the tree here, in which case we add the data from this level onto theirs.

3. The tree below is narrower, for instance an only child.

Value

A list containing the elements to plot the Pedigree. It contains a set of matrices along with the spouselist matrix. The latter has marriages removed as they are processed.

• n : A vector giving the number of subjects on each horizonal level of the plot

• nid : A matrix with one row for each level, giving the numeric id of each subject plotted. (A value of 17 means the 17th subject in the Pedigree).

• pos : A matrix giving the horizontal position of each plot point

• fam : A matrix giving the family id of each plot point. A value of 3 would mean that the two subjects in positions 3 and 4, in the row above, are this subject's parents.

• spouselist : Spouse matrix with anchors informations

See Also

align()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
align(ped)

Description

Second of the four co-routines which takes a collection of siblings, grows the tree for each, and appends them side by side into a single tree.

Usage

alignped2(idx, dadx, momx, level, horder, packed, spouselist)

Arguments

Details

The input arguments are the same as those to alignped1() with the exception that idx will be a vector. This routine does nothing to the spouselist matrix, but needs to pass it down the tree and back since one of the routines called by alignped2() might change the matrix.

The code below has one non-obvious special case. Suppose that two sibs marry. When the first sib is processed by alignped1 then both partners (and any children) will be added to the rval structure below. When the second sib is processed they will come back as a 1 element tree (the marriage will no longer be on the spouselist), which should be added onto rval. The rule thus is to not add any 1 element tree whose value (which must be idx[i] is already in the rval structure for this level.

Value

A list containing the elements to plot the Pedigree. It contains a set of matrices along with the spouselist matrix. The latter has marriages removed as they are processed.

• n : A vector giving the number of subjects on each horizonal level of the plot

• nid : A matrix with one row for each level, giving the numeric id of each subject plotted. (A value of 17 means the 17th subject in the Pedigree).

• pos : A matrix giving the horizontal position of each plot point

• fam : A matrix giving the family id of each plot point. A value of 3 would mean that the two subjects in positions 3 and 4, in the row above, are this subject's parents.

• spouselist : Spouse matrix with anchors informations

See Also

align()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
align(ped)

Description

Third of the four co-routines to merges two pedigree trees which are side by side into a single object.

Usage

alignped3(alt1, alt2, packed, space = 1)

Arguments

Details

The primary special case is when the rightmost person in the left tree is the same as the leftmost person in the right tree; we need not plot two copies of the same person side by side. (When initializing the output structures do not worry about this, there is no harm if they are a column bigger than finally needed.) Beyond that the work is simple book keeping.

1. Slide:

For the unpacked case, which is the traditional way to draw a Pedigree when we can assume the paper is infinitely wide, all parents are centered over their children. In this case we think if the two trees to be merged as solid blocks. On input they both have a left margin of 0. Compute how far over we have to slide the right tree.

2. Merge:

Now merge the two trees. Start at the top level and work down.

Value

A list containing the elements to plot the Pedigree. It contains a set of matrices along with the spouselist matrix. The latter has marriages removed as they are processed.

• n : A vector giving the number of subjects on each horizonal level of the plot

• nid : A matrix with one row for each level, giving the numeric id of each subject plotted. (A value of 17 means the 17th subject in the Pedigree).

• pos : A matrix giving the horizontal position of each plot point

• fam : A matrix giving the family id of each plot point. A value of 3 would mean that the two subjects in positions 3 and 4, in the row above, are this subject's parents.

• spouselist : Spouse matrix with anchors informations

See Also

align()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
align(ped)

Description

Last routines which attempts to line up children under parents and put spouses and siblings "close" to each other, to the extent possible within the constraints of page width.

Usage

alignped4(rval, spouse, level, width, align)

Arguments

Details

The alignped4() routine is the final step of alignment. The current code does necessary setup and then calls the quadprog::solve.QP() function.

There are two important parameters for the function:

1. The maximum width specified. The smallest possible width is the maximum number of subjects on a line, if the user suggestion is too low it is increased to that 1 + that amount (to give just a little wiggle room).

2. The align vector of 2 alignment parameters a and b. For each set of siblings x with parents at p_1 and p_2 the alignment penalty is :

   $(1/k^a)\sum{i=1}{k} (x_i - (p_1 + p_2)^2$

   where k is the number of siblings in the set.

Using the fact that when a = 1 :

$\sum(x_i-c)^2 = \sum(x_i-\mu)^2 + k(c-\mu)^2$

then moving a sibship with k sibs one unit to the left or right of optimal will incur the same cost as moving one with only 1 or two sibs out of place.

If a = 0 then large sibships are harder to move than small ones, with the default value a = 1.5 they are slightly easier to move than small ones. The rationale for the default is as long as the parents are somewhere between the first and last siblings the result looks fairly good, so we are more flexible with the spacing of a large family. By tethering all the sibs to a single spot they tend are kept close to each other.

The alignment penalty for spouses is $b(x_1 - x_2)^2$, which tends to keep them together. The size of b controls the relative importance of sib-parent and spouse-spouse closeness.

1. We start by adding in these penalties. The total number of parameters in the alignment problem (what we hand to quadprog) is the set of sum(n) positions. A work array myid keeps track of the parameter number for each position so that it is easy to find. There is one extra penalty added at the end. Because the penalty amount would be the same if all the final positions were shifted by a constant, the penalty matrix will not be positive definite; solve.QP() does not like this. We add a tiny amount of leftward pull to the widest line.

2. If there are k subjects on a line there will be k+1 constraints for that line. The first point must be $\ge 0$, each subesquent one must be at least 1 unit to the right, and the final point must be $\le$ the max width.

Value

The updated position matrix

See Also

align()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
align(ped)

Description

Anchor variable to ordered factor

Usage

anchor_to_factor(anchor)

Arguments

Value

An ordered factor vector containing the transformed variable "either" < "left" < "right"

Examples

anchor_to_factor(c(1, 2, 0, "left", "right", "either"))

Description

Compute an initial guess for the alignment of a Pedigree

Usage

## S4 method for signature 'Pedigree'
auto_hint(obj, hints = NULL, packed = TRUE, align = FALSE, reset = FALSE)

Arguments

Details

A Pedigree structure can contain a Hints object which helps to reorder the Pedigree (e.g. left-to-right order of children within family) so as to plot with minimal distortion. This routine is used to create an initial version of the hints. They can then be modified if desired.

This routine would not normally be called by a user. It moves children within families, so that marriages are on the "edge" of a set children, closest to the spouse. For pedigrees that have only a single connection between two families this simple-minded approach works surprisingly well. For more complex structures hand-tuning of the hints may be required.

When auto_hint() is called with a a vector of numbers as the hints argument, the values for the founder females are used to order the founder families left to right across the plot. The values within a sibship are used as the preliminary order of siblings within a family; this may be changed to move one of them to the edge so as to match up with a spouse. The actual values in the vector are not important, only their order.

Value

The initial Hints object.

See Also

align(), best_hint(), Hints

Examples

data(sampleped)
ped <- Pedigree(sampleped[sampleped$famid == 1, ])
auto_hint(ped)

Description

When computer time is cheap, use this routine to get a best Pedigree alignment. This routine will try all possible founder orders, and return the one with the least stress.

Usage

## S4 method for signature 'Pedigree'
best_hint(obj, wt = c(1000, 10, 1), tolerance = 0)

Arguments

Details

The auto_hint() routine will rearrange sibling order, but not founder order. This calls auto_hint() with every possible founder order, and finds that plot with the least "stress". The stress is computed as a weighted sum of three error measures:

• nbArcs The number of duplicate individuals in the plot

• lgArcs The sum of the absolute values of the differences in the positions of duplicate individuals

• lgParentsChilds The sum of the absolute values of the differences between the center of the children and the parents

$stress = wt[1] * nbArcs + wt[2] * lgArcs + wt[3] * lgParentsChilds$

If during the search, a plot is found with a stress level less than tolerance, the search is terminated.

Value

The best Hints object out of all the permutations

See Also

auto_hint(), align()

Examples

data(sampleped)
ped <- Pedigree(sampleped[sampleped$famid == 1,])
best_hint(ped)

Description

Utility function used in the shrink() function to calculate the bit size of a Pedigree.

Usage

## S4 method for signature 'character_OR_integer'
bit_size(obj, momid, missid = NA_character_)

## S4 method for signature 'Pedigree'
bit_size(obj)

## S4 method for signature 'Ped'
bit_size(obj)

Arguments

Details

The bit size of a Pedigree is defined as :

$2 \times NbNonFounders - NbFounders$

Where NbNonFounders is the number of non founders in the Pedigree (i.e. individuals with identified parents) and NbFounders is the number of founders in the Pedigree (i.e. individuals without identified parents).

Value

A list with the following components:

• bit_size The bit size of the Pedigree

• nFounder The number of founders in the Pedigree

• nNonFounder The number of non founders in the Pedigree

See Also

shrink()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
bit_size(ped)

Description

Create a list of x and y coordinates for a circle with a given number of slices.

Usage

circfun(nslice, n = 50)

Arguments

Value

A list of x and y coordinates per slice.

Examples

circfun(1)
circfun(1, 10)
circfun(4, 50)

Description

Draw arcs

Usage

draw_arc(x0, y0, x1, y1, p, ggplot_gen = FALSE, lwd = 1, col = "black")

Arguments

Value

Plot the arcs to the current device or add it to a ggplot object

Description

Draw a polygon

Usage

draw_polygon(
  x,
  y,
  p,
  ggplot_gen = FALSE,
  fill = "grey",
  border = NULL,
  density = NULL,
  angle = 45
)

Arguments

Value

Plot the polygon to the current device or add it to a ggplot object

Description

Draw segments

Usage

draw_segment(
  x0,
  y0,
  x1,
  y1,
  p,
  ggplot_gen,
  col = par("fg"),
  lwd = par("lwd"),
  lty = par("lty")
)

Arguments

Value

Plot the segments to the current device or add it to a ggplot object

Description

Draw texts

Usage

draw_text(
  x,
  y,
  label,
  p,
  ggplot_gen = FALSE,
  cex = 1,
  col = NULL,
  adjx = 0,
  adjy = 0
)

Arguments

Value

Plot the text to the current device or add it to a ggplot object

Description

Find the duplicate pairs of a subject

Usage

duporder(idlist, plist, lev, obj)

Arguments

Details

This routine is used by auto_hint(). It finds the duplicate pairs of a subject and returns them in the order they should be plotted.

Value

A matrix of duplicate pairs

See Also

auto_hint()

Description

Exclude any founders who are not parents.

Usage

exclude_stray_marryin(id, dadid, momid)

Arguments

Value

Returns a data frame of subject identifiers and their parents. The data frame is trimmed of any founders who are not parents.

See Also

shrink()

Description

Exclude any unavailable founders.

Usage

exclude_unavail_founders(id, dadid, momid, avail, missid = NA_character_)

Arguments

Value

Returns a list with the following components:

• n_trimmed Number of trimmed individuals

• id_trimmed Vector of IDs of trimmed individuals

• id Vector of subject identifiers

• dadid Vector of father identifiers

• momid Vector of mother identifiers

See Also

shrink()

Description

Finds one subject from among available non-parents with indicated affection status.

Usage

## S4 method for signature 'Ped'
find_avail_affected(obj, avail = NULL, affected = NULL, affstatus = NA)

## S4 method for signature 'Pedigree'
find_avail_affected(obj, avail = NULL, affected = NULL, affstatus = NA)

Arguments

Details

When used within shrink(), this function is called with the first affected indicator, if the affected item in the Pedigree is a matrix of multiple affected indicators.

If avail or affected is null, then the function will use the corresponding Ped accessor.

Value

A list is returned with the following components

• ped The new Ped object

• newAvail Vector of availability status of trimmed individuals

• idTrimmed Vector of IDs of trimmed individuals

• isTrimmed logical value indicating whether Ped object has been trimmed

• bit_size Bit size of the trimmed Ped

See Also

shrink()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
find_avail_affected(ped, affstatus = 1)

Description

Finds subjects from among available non-parents with all affection equal to 0.

Usage

## S4 method for signature 'Ped'
find_avail_noninform(obj, avail = NULL, affected = NULL)

## S4 method for signature 'Pedigree'
find_avail_noninform(obj, avail = NULL, affected = NULL)

Arguments

Details

Identify subjects to remove from a Pedigree who are available but non-informative (unaffected). This is the second step to remove subjects in shrink() if the Pedigree does not meet the desired bit size.

If avail or affected is null, then the function will use the corresponding Ped accessor.

Value

Vector of subject ids who can be removed by having lowest informativeness.

See Also

shrink()

Examples

data(sampleped)
ped <- Pedigree(sampleped)
find_avail_noninform(ped)

Description

Find the identifiers of subjects in a Pedigree iteratively, as anyone who is not available and does not have an available descendant by successively removing unavailable terminal nodes.

Usage

## S4 method for signature 'Ped'
find_unavailable(obj, avail = NULL)

## S4 method for signature 'Pedigree'
find_unavailable(obj, avail = NULL)

Arguments

Details

If avail is null, then the function will use the corresponding Ped accessor.

Originally written as pedTrim by Steve Iturria, modified by Dan Schaid 2007, and now split into the two separate functions: find_unavailable(), and trim() to do the tasks separately. find_unavailable() calls exclude_stray_marryin() to find stray available marry-ins who are isolated after trimming their unavailable offspring, and exclude_unavail_founders(). If the subject ids are character, make sure none of the characters in the ids is a colon (":"), which is a special character used to concatenate and split subjects within the utility. The trim() functions is now replaced by the subset() function.

Value

Returns a vector of subject ids for who can be removed.

Side Effects

Relation matrix from subsetting is trimmed of any special relations that include the subjects to trim.

See Also

shrink()

Examples

data(sampleped)
ped1 <- Pedigree(sampleped[sampleped$famid == "1",])
find_unavailable(ped1)

Description

Find the siblings of a subject

Usage

findsibs(idpos, plist, lev)

Arguments

Details

This routine is used by auto_hint(). It finds the siblings of a subject.

Value

The positions of the siblings

See Also

auto_hint()

Description

Find the spouse of a subject

Usage

findspouse(idpos, plist, lev, obj)

Arguments

Details

This routine is used by auto_hint(). It finds the spouse of a subject.

Value

The position of the spouse

See Also

auto_hint()

Description

Fix the sex of parents, add parents that are missing from the data. Can be used with a dataframe or a vector of the different individuals informations.

Usage

## S4 method for signature 'character'
fix_parents(obj, dadid, momid, sex, famid = NULL, missid = NA_character_)

## S4 method for signature 'data.frame'
fix_parents(obj, delete = FALSE, filter = NULL, missid = NA_character_)

Arguments

Details

First look to add parents whose ids are given in momid/dadid. Second, fix sex of parents. Last look to add second parent for children for whom only one parent id is given. If a famid vector is given the family id will be added to the ids of all individuals (id, dadid, momid) separated by an underscore before proceeding.

Special case for dataframe

Check for presence of both parents id in the id field. If not both presence behaviour depend of delete parameter

• If TRUE then use fix_parents function and merge back the other fields in the dataframe then set availability to O for non available parents.

• If FALSE then delete the id of missing parents

Value

A data.frame with id, dadid, momid, sex as columns with the relationships fixed.

Author(s)

Jason Sinnwell

Examples

test1char <- data.frame(
  id = paste('fam', 101:111, sep = ''),
  sex = c('male', 'female')[c(1, 2, 1, 2, 1, 1, 2, 2, 1, 2, 1)],
  father = c(
    0, 0, 'fam101', 'fam101', 'fam101', 0, 0,
    'fam106', 'fam106', 'fam106', 'fam109'
  ),
  mother = c(
    0, 0, 'fam102', 'fam102', 'fam102', 0, 0,
    'fam107', 'fam107', 'fam107', 'fam112'
  )
)
test1newmom <- with(test1char, fix_parents(id, father, mother,
  sex,
  missid = NA_character_
))
Pedigree(test1newmom)

Description

Perform transformation uppon a vector given as the one containing the affection status to obtain an affected binary state.

Usage

generate_aff_inds(
  values,
  mods_aff = NULL,
  threshold = NULL,
  sup_thres_aff = NULL
)

Arguments

Details

This function helps to configure a binary state from a character or numeric variable.

If the variable is a character or a factor:

In this case the affected state will depend on the modality provided as an affected status. All individuals with a value corresponding to one of the element in the vector mods_aff will be considered as affected.

If the variable is numeric:

In this case the affected state will be TRUE if the value of the individual is above the threshold if sup_thres_aff is TRUE and FALSE otherwise.

Value

A dataframe with the affected column processed accordingly. The different columns are:

• mods: The different modalities of the column

• labels: The labels of the different modalities

• affected: The column processed to have only TRUE/FALSE values

Author(s)

Louis Le Nézet

Examples

generate_aff_inds(c(1, 2, 3, 4, 5), threshold = 3, sup_thres_aff = TRUE)
generate_aff_inds(c("A", "B", "C", "A", "V", "B"), mods_aff = c("A", "B"))

Description

Perform transformation uppon a vector given as the one containing the availability status to compute the border color. The vector given will be transformed using the vect_to_binary() function.

Usage

generate_border(values, colors_avail = c("green", "black"))

Arguments

Value

A list of three elements

• mods : The processed values column as a numeric factor

• avail : A logical vector indicating if the individual is available

• sc_bord : A dataframe containing the description of each modality of the scale

Examples

generate_border(c(1, 0, 1, 0, NA, 1, 0, 1, 0, NA))

Description

Perform transformation uppon a dataframe given to compute the colors for the filling and the border of the individuals based on the affection and availability status.

Usage

## S4 method for signature 'character'
generate_colors(
  obj,
  avail,
  mods_aff = NULL,
  keep_full_scale = FALSE,
  colors_aff = c("yellow2", "red"),
  colors_unaff = c("white", "steelblue4"),
  colors_avail = c("green", "black")
)

## S4 method for signature 'numeric'
generate_colors(
  obj,
  avail,
  threshold = 0.5,
  sup_thres_aff = TRUE,
  keep_full_scale = FALSE,
  breaks = 3,
  colors_aff = c("yellow2", "red"),
  colors_unaff = c("white", "steelblue4"),
  colors_avail = c("green", "black")
)

## S4 method for signature 'Pedigree'
generate_colors(
  obj,
  col_aff = "affected",
  add_to_scale = TRUE,
  col_avail = "avail",
  mods_aff = NULL,
  threshold = 0.5,
  sup_thres_aff = TRUE,
  keep_full_scale = FALSE,
  breaks = 3,
  colors_aff = c("yellow2", "red"),
  colors_unaff = c("white", "steelblue4"),
  colors_avail = c("green", "black"),
  reset = TRUE
)

Arguments

Details

The colors will be set using the generate_fill() and the generate_border() functions respectively for the filling and the border.

Value

When used with a vector

A list of two elements

• The list containing the filling colors processed and their description

• The list containing the border colors processed and their description

When used with a Pedigree object

The Pedigree object with the affected and avail columns processed accordingly as well as the scales slot updated.

Examples

generate_colors(
    c("A", "B", "A", "B", NA, "A", "B", "A", "B", NA),
    c(1, 0, 1, 0, NA, 1, 0, 1, 0, NA),
    mods_aff = "A",
)

generate_colors(
    c(10, 0, 5, 7, NA, 6, 2, 1, 3, NA),
    c(1, 0, 1, 0, NA, 1, 0, 1, 0, NA),
    threshold = 3, keep_full_scale = TRUE
)
data("sampleped")
ped <- Pedigree(sampleped)
ped <- generate_colors(ped, "affected", add_to_scale=FALSE)
scales(ped)

Description

Perform transformation uppon a column given as the one containing affection status to compute the filling color.

Usage

generate_fill(
  values,
  affected,
  labels,
  keep_full_scale = FALSE,
  breaks = 3,
  colors_aff = c("yellow2", "red"),
  colors_unaff = c("white", "steelblue4")
)

Arguments

Details

The colors will be set using the grDevices::colorRampPalette() function with the colors given as parameters.

The colors will be set as follow:

• If keep_full_scale is FALSE: Then the affected individuals will get the first color of the colors_aff vector and the unaffected individuals will get the first color of the colors_unaff vector.

• If keep_full_scale is TRUE:

  • If values isn't numeric: Each levels of the affected values vector will get it's own color from the colors_aff vector using the grDevices::colorRampPalette() and the same will be done for the unaffected individuals using the colors_unaff.

  • If values is numeric: The mean of the affected individuals will be compared to the mean of the unaffected individuals and the colors will be set up such as the color gradient follow the direction of the affection.

Value

A list of three elements

• mods : The processed values column as a numeric factor

• affected : A logical vector indicating if the individual is affected

• sc_fill : A dataframe containing the description of each modality of the scale

Examples

aff <- generate_aff_inds(seq_len(5), threshold = 3, sup_thres_aff = TRUE)
generate_fill(seq_len(5), aff$affected, aff$labels)
generate_fill(seq_len(5), aff$affected, aff$labels, keep_full_scale = TRUE)

Description

Get twin relationships

Usage

get_twin_rel(obj)

Arguments

Details

This routine function determine the twin relationships in a Pedigree. It determine the order of the twins in the Pedigree. It is used by auto_hint().

Value

A list containing components

1. twinset the set of twins

2. twinrel the twins relationships

3. twinord the order of the twins

See Also

auto_hint()

Description

The hints are used to specify the order of the individuals in the pedigree and to specify the order of the spouses.

Constructor :

You either need to provide horder or spouse in the dedicated parameters (together or separately), or inside a list.

Usage

Hints(horder, spouse)

## S4 method for signature 'list,missing_OR_NULL'
Hints(horder, spouse)

## S4 method for signature 'numeric,data.frame'
Hints(horder, spouse)

## S4 method for signature 'numeric,missing_OR_NULL'
Hints(horder, spouse)

Arguments

Value

A Hints object.

Slots

horder

A numeric named vector with one element per subject in the Pedigree. It determines the relative horizontal order of subjects within a sibship, as well as the relative order of processing for the founder couples. (For this latter, the female founders are ordered as though they were sisters).

spouse

A data.frame with one row per hinted marriage, usually only a few marriages in a Pedigree will need an added hint, for instance reverse the plot order of a husband/wife pair. Each row contains the identifiers of the left spouse, the right hand spouse, and the anchor (i.e : 1 = left, 2 = right, 0 = either).

Accessors

• horder(x) : Get the horder vector

• horder(x) <- value : Set the horder vector

• spouse(x) : Get the spouse data.frame

• spouse(x) <- value : Set the spouse data.frame

Generics

• as.list(x): Convert a Hints object to a list

• subset(x, i, keep = TRUE): Subset a Hints object based on the individuals identifiers given.

  • i : A vector of individuals identifiers to keep.

  • keep : A logical value indicating if the individuals should be kept or deleted.

See Also

Pedigree()

Examples

Hints(
  list(
      horder = c("1" = 1, "2" = 2, "3" = 3),
      spouse = data.frame(
          idl = c("1", "2"),
          idr = c("2", "3"),
          anchor = c(1, 2)
     )
  )
)

Hints(
  horder = c("1" = 1, "2" = 2, "3" = 3),
  spouse = data.frame(
      idl = c("1", "2"),
      idr = c("2", "3"),
      anchor = c(1, 2)
  )
)

Hints(
  horder = c("1" = 1, "2" = 2, "3" = 3)
)

Description

Transform identity by descent (IBD) matrix data from the form produced by external programs such as SOLAR into the compact form used by the coxme and lmekin routines.

Usage

ibd_matrix(id1, id2, ibd, idmap, diagonal)

Arguments

Details

The IBD matrix for a set of n subjects will be an n by n symmetric matrix whose i,j element is the contains, for some given genetic location, a 0/1 indicator of whether 0, 1/2 or 2/2 of the alleles for i and j are identical by descent. Fractional values occur if the IBD fraction must be imputed. The diagonal will be 1. Since a large fraction of the values will be zero, programs such as Solar return a data set containing only the non-zero elements. As well, Solar will have renumbered the subjects as seq_len(n) in such a way that families are grouped together in the matrix; a separate index file contains the mapping between this new id and the original one. The final matrix should be labeled with the original identifiers.

Value

a sparse matrix of class dsCMatrix. This is the same form used for kinship matrices.

See Also

kinship()

Description

Select the ids of the informative individuals.

Usage

## S4 method for signature 'character_OR_integer'
is_informative(obj, avail, affected, informative = "AvAf")

## S4 method for signature 'Pedigree'
is_informative(obj, col_aff = NULL, informative = "AvAf", reset = FALSE)

Arguments

Details

Depending on the informative parameter, the function will extract the ids of the informative individuals. In the case of a numeric vector, the function will return the same vector. In the case of a boolean, the function will return the ids of the individuals if TRUE, NA otherwise. In the case of a string, the function will return the ids of the corresponding informative individuals based on the avail and affected columns.

Value

When obj is a vector

A vector of individuals informative identifiers.

When obj is a Pedigree

The Pedigree object with its isinf slot updated.

Examples

is_informative(c("A", "B", "C", "D", "E"), informative = c("A", "B"))
is_informative(c("A", "B", "C", "D", "E"), informative = c(1, 2))
is_informative(c("A", "B", "C", "D", "E"), informative = c("A", "B"))
is_informative(c("A", "B", "C", "D", "E"), avail = c(1, 0, 0, 1, 1),
  affected = c(0, 1, 0, 1, 1), informative = "AvAf")
is_informative(c("A", "B", "C", "D", "E"), avail = c(1, 0, 0, 1, 1),
  affected = c(0, 1, 0, 1, 1), informative = "AvOrAf")
is_informative(c("A", "B", "C", "D", "E"),
     informative = c(TRUE, FALSE, TRUE, FALSE, TRUE))

data("sampleped")
ped <- Pedigree(sampleped)
ped <- is_informative(ped, col_aff = "affection_mods")
isinf(ped(ped))

Description

Computes the depth of each subject in the Pedigree.

Usage

kindepth(obj, ...)

## S4 method for signature 'character_OR_integer'
kindepth(obj, dadid, momid, align_parents = FALSE)

## S4 method for signature 'Pedigree'
kindepth(obj, align_parents = FALSE)

## S4 method for signature 'Ped'
kindepth(obj, align_parents = FALSE)

Arguments

Details

Mark each person as to their depth in a Pedigree; 0 for a founder, otherwise :

$depth = 1 + \max(fatherDepth, motherDepth)$

In the case of an inbred Pedigree a perfect alignment may not exist.

Value

An integer vector containing the depth for each subject

Author(s)

Terry Therneau, updated by Louis Le Nézet

See Also

align()

Examples

kindepth(
     c("A", "B", "C", "D", "E"),
     c("C", "D", "0", "0", "0"),
     c("E", "E", "0", "0", "0")
)
data(sampleped)
ped1 <- Pedigree(sampleped[sampleped$famid == "1",])
kindepth(ped1)

Description

Compute the kinship matrix for a set of related autosomal subjects. The function is generic, and can accept a Pedigree, a Ped or a vector as the first argument.

Usage

## S4 method for signature 'Ped'
kinship(obj, chrtype = "autosome")

## S4 method for signature 'character'
kinship(obj, dadid, momid, sex, chrtype = "autosome")

## S4 method for signature 'Pedigree'
kinship(obj, chrtype = "autosome")

Arguments

Details

The function will usually be called with a Pedigree. The call with a Ped or a vector is provided for backwards compatibility with an earlier release of the library that was less capable. Note that when using with a Ped or a vector, any information on twins is not available to the function.

When called with a Pedigree, the routine will create a block-diagonal-symmetric sparse matrix object of class dsCMatrix. Since the ⁠[i, j]⁠ value of the result is 0 for any two unrelated individuals i and j and a Matrix utilizes sparse representation, the resulting object is often orders of magnitude smaller than an ordinary matrix.

Two genes G1 and G2 are identical by descent (IBD) if they are both physical copies of the same ancestral gene; two genes are identical by state if they represent the same allele. So the brown eye gene that I inherited from my mother is IBD with hers; the same gene in an unrelated individual is not.

The kinship coefficient between two subjects is the probability that a randomly selected allele from a locus will be IBD between them. It is obviously 0 between unrelated individuals. For an autosomal site and no inbreeding it will be 0.5 for an individual with themselves, .25 between mother and child, .125 between an uncle and neice, etc.

The computation is based on a recursive algorithm described in Lange, which assumes that the founder alleles are all independent.

Value

When obj is a vector

A matrix of kinship coefficients.

When obj is a Pedigree

A matrix of kinship coefficients ordered by families present in the Pedigree object.

References

K Lange, Mathematical and Statistical Methods for Genetic Analysis, Springer-Verlag, New York, 1997.

See Also

make_famid(), kindepth()

Examples

kinship(c("A", "B", "C", "D", "E"), c("C", "D", "0", "0", "0"),
    c("E", "E", "0", "0", "0"), sex = c(1, 2, 1, 2, 1))
kinship(c("A", "B", "C", "D", "E"), c("C", "D", "0", "0", "0"),
    c("E", "E", "0", "0", "0"), sex = c(1, 2, 1, 2, 1),
   chrtype = "x"
)

data(sampleped)
ped <- Pedigree(sampleped)
kinship(ped)

Description

Construct a family identifier from pedigree information

Usage

## S4 method for signature 'character'
make_famid(obj, dadid, momid)

## S4 method for signature 'Pedigree'
make_famid(obj)

Arguments

Details

Create a vector of length n, giving the family 'tree' number of each subject. If the Pedigree is totally connected, then everyone will end up in tree 1, otherwise the tree numbers represent the disconnected subfamilies. Singleton subjects give a zero for family number.

Value

When used with a character vector

An integer vector giving family groupings

When used with a Pedigree object

An updated Pedigree object with the family id added and with all ids updated

See Also

kinship()

Examples

make_famid(
     c("A", "B", "C", "D", "E", "F"),
     c("C", "D", "0", "0", "0", "0"),
     c("E", "E", "0", "0", "0", "0")
)

data(sampleped)
ped1 <- Pedigree(sampleped[,-1])
make_famid(ped1)

Description

Compute the minimum distance between the informative individuals and all the others. This distance is a transformation of the maximum kinship degree between the informative individuals and all the others. This transformation is done by taking the log2 of the inverse of the maximum kinship degree.

$minDist = log2(1 / \max(kinship))$

Therefore, the minimum distance is 0 when the maximum kinship is 1 and is infinite when the maximum kinship is 0. For siblings, the kinship value is 0.5 and the minimum distance is 1. Each time the kinship degree is divided by 2, the minimum distance is increased by 1.

Usage

## S4 method for signature 'character'
min_dist_inf(obj, dadid, momid, sex, id_inf)

## S4 method for signature 'Pedigree'
min_dist_inf(obj, col_aff = NULL, informative = "AvAf", reset = FALSE, ...)

## S4 method for signature 'Ped'
min_dist_inf(obj, informative = "AvAf", reset = FALSE)

Arguments

Value

When obj is a vector

A vector of the minimum distance between the informative individuals and all the others corresponding to the order of the individuals in the obj vector.

When obj is a Pedigree

The Pedigree object with a new slot named 'kin' containing the minimum distance between each individuals and the informative individuals. The isinf slot is also updated with the informative individuals.

See Also

kinship()

Examples

min_dist_inf(
     c("A", "B", "C", "D", "E"),
     c("C", "D", "0", "0", "0"),
     c("E", "E", "0", "0", "0"),
     sex = c(1, 2, 1, 2, 1),
     id_inf = c("D", "E")
)

data(sampleped)
ped <- Pedigree(sampleped)
kin(ped(min_dist_inf(ped, col_aff = "affection_mods")))

Description

Data from the Minnesota Breast Cancer Family Study. This contains extended pedigrees from 426 families, each identified by a single proband in 1945-1952, with follow up for incident breast cancer.

Usage

data(minnbreast)

Format

A data frame with 28081 observations, one line per subject, on the following 14 variables.

• id : Subject identifier

• proband : If 1, this subject is one of the original 426 probands

• fatherid : Identifier of the father, if the father is part of the data set; zero otherwise

• motherid : Identifier of the mother, if the mother is part of the data set; zero otherwise

• famid : Family identifier

• endage : Age at last follow-up or incident cancer

• cancer : 1 = breast cancer (females) or prostate cancer (males), 0 = censored

• yob : Year of birth

• education : Amount of education: 1-8 years, 9-12 years, high school graduate, vocational education beyond high school, some college but did not graduate, college graduate, post-graduate education, refused to answer on the questionnaire

• marstat : Marital status: married, living with someone in a marriage-like relationship, separated or divorced, widowed, never married, refused to answer the questionaire

• everpreg : Ever pregnant at the time of baseline survey

• parity : Number of births

• nbreast : Number of breast biopsies

• sex : M or F

• bcpc : Part of one of the families in the breast / prostate cancer substudy: 0 = no, 1 = yes. Note that subjects who were recruited to the overall study after the date of the BP substudy are coded as zero.

Details

The original study was conducted by Dr. Elving Anderson at the Dight Institute for Human Genetics at the University of Minnesota. From 1944 to 1952, 544 sequential breast cancer cases seen at the University Hospital were enrolled, and information gathered on parents, siblings, offspring, aunts / uncles, and grandparents with the goal of understanding possible familial aspects of brest cancer. In 1991 the study was resurrected by Dr Tom Sellers.

Of the original 544 he excluded 58 prevalent cases, along with another 19 who had less than 2 living relatives at the time of Dr Anderson's survey. Of the remaining 462 families 10 had no living members, 23 could not be located and 8 refused, leaving 426 families on whom updated pedigrees were obtained.

This gave a study with 13351 males and 12699 females (5183 marry-ins). Primary questions were the relationship of early life exposures, breast density, and pharmacogenomics on incident breast cancer risk. For a subset of the families data was gathered on prostate cancer risk for male subjects via questionnaires sent to men over 40. Other than this, data items other than parentage are limited to the female subjects. In 2003 a second phase of the study was instituted. The pedigrees were further extended to the numbers found in this data set, and further data gathered by questionnaire.

References

Epidemiologic and genetic follow-up study of 544 Minnesota breast cancer families: design and methods. Sellers TA, Anderson VE, Potter JD, Bartow SA, Chen PL, Everson L, King RA, Kuni CC, Kushi LH, McGovern PG, et al. Genetic Epidemiology, 1995; 12(4):417-29.

Evaluation of familial clustering of breast and prostate cancer in the Minnesota Breast Cancer Family Study. Grabrick DM, Cerhan JR, Vierkant RA, Therneau TM, Cheville JC, Tindall DJ, Sellers TA. Cancer Detect Prev. 2003; 27(1):30-6.

Risk of breast cancer with oral contraceptive use in women with a family history of breast cancer. Grabrick DM, Hartmann LC, Cerhan JR, Vierkant RA, Therneau TM, Vachon CM, Olson JE, Couch FJ, Anderson KE, Pankratz VS, Sellers TA. JAMA. 2000; 284(14):1791-8.

Examples

data(minnbreast)
breastped <- Pedigree(minnbreast,
   cols_ren_ped = list(
       "indId" = "id", "fatherId" = "fatherid",
       "motherId" = "motherid", "gender" = "sex", "family" = "famid"
   ), missid = "0", col_aff = "cancer"
)
summary(breastped)
scales(breastped)
#plot family 8, proband is solid, slash for cancers
#plot(breastped[famid(breastped) == "8"])

Description

Normalise dataframe for a Ped object

Usage

norm_ped(
  ped_df,
  na_strings = c("NA", ""),
  missid = NA_character_,
  try_num = FALSE
)

Arguments

- `indID` individual identifiers -> `id`
- `fatherId` biological fathers identifiers -> `dadid`
- `motherId` biological mothers identifiers -> `momdid`
- `gender` sex of the individual -> `sex`
- `family` family identifiers -> `famid`

- `sterilisation` status -> `steril`
- `available` status -> `avail`
- `vitalStatus`, is the individual dead -> `status`
- `affection` status -> `affected`

Details

Normalise a dataframe and check for columns correspondance to be able to use it as an input to create a Ped object. Multiple test are done and errors are checked. Sex is calculated based on the gender column.

The steril column need to be a boolean either TRUE, FALSE or 'NA'. Will be considered available any individual with no 'NA' values in the available column. Duplicated indId will nullify the relationship of the individual. All individuals with errors will be remove from the dataframe and will be transfered to the error dataframe.

A number of checks are done to ensure the dataframe is correct:

On identifiers:

• All ids (id, dadid, momid, famid) are not empty (⁠!= ""⁠)

• All id are unique (no duplicated)

• All dadid and momid are unique in the id column (no duplicated)

• id is not the same as dadid or momid

• Either have both parents or none

On sex

• All sex code are either male, female, terminated or unknown.

• No parents are steril

• All fathers are male

• All mothers are female

Value

A dataframe with different variable correctly standardized and with the errors identified in the error column

See Also

Ped(), Ped, Pedigree()

Examples

df <- data.frame(
    indId = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10),
    fatherId = c("A", 0, 1, 3, 0, 4, 1, 0, 6, 6),
    motherId = c(0, 0, 2, 2, 0, 5, 2, 0, 8, 8),
    gender = c(1, 2, "m", "man", "f", "male", "m", "m", "f", "f"),
    available = c("A", "1", 0, NA, 1, 0, 1, 0, 1, 0),
    famid = c(1, 1, 1, 1, 1, 1, 1, 2, 2, 2),
    sterilisation = c("TRUE", "FALSE", TRUE, FALSE, 1, 0, 1, 0, 1, "TRUE"),
    vitalStatus = c("TRUE", "FALSE", TRUE, FALSE, 1, 0, 1, 0, 1, 0),
    affection = c("TRUE", "FALSE", TRUE, FALSE, 1, 0, 1, 0, 1, 0)
)
tryCatch(
     norm_ped(df),
     error = function(e) print(e)
)

Description

Normalise a dataframe and check for columns correspondance to be able to use it as an input to create a Ped object.

Usage

norm_rel(rel_df, na_strings = c("NA", ""), missid = NA_character_)

Arguments

Details

The famid column, if provided, will be merged to the ids field separated by an underscore using the upd_famid_id() function. The code column will be transformed with the rel_code_to_factor(). Multiple test are done and errors are checked.

A number of checks are done to ensure the dataframe is correct:

On identifiers:

• All ids (id1, id2) are not empty (⁠!= ""⁠)

• id1 and id2 are not the same

On code

• All code are recognised as either "MZ twin", "DZ twin", "UZ twin" or "Spouse"

Value

A dataframe with the errors identified

Examples

df <- data.frame(
   id1 = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10),
   id2 = c(2, 3, 4, 5, 6, 7, 8, 9, 10, 1),
   code = c("MZ twin", "DZ twin", "UZ twin", "Spouse", 1, 2,
      3, 4, "MzTwin", "sp oUse"),
   famid = c(1, 1, 1, 1, 1, 1, 1, 2, 2, 2)
)
norm_rel(df)

Description

Compute the number of childs per individual

Usage

## S4 method for signature 'character_OR_integer'
num_child(obj, dadid, momid, rel_df = NULL, missid = NA_character_)

## S4 method for signature 'Pedigree'
num_child(obj, reset = FALSE)

Arguments

Details

Compute the number of direct child but also the number of indirect child given by the ones related with the linked spouses. If a relation ship dataframe is given, then even if no children is present between 2 spouses, the indirect childs will still be added.

Value

When obj is a vector

A dataframe with the columns num_child_dir, num_child_ind and num_child_tot giving respectively the direct, indirect and total number of child.

When obj is a Pedigree object

An updated Pedigree object with the columns num_child_dir, num_child_ind and num_child_tot added to the Pedigree ped slot.

Examples

num_child(
  obj = c("1", "2", "3", "4", "5", "6", "7", "8", "9", "10"),
  dadid = c("3", "3", "6", "8", "0", "0", "0", "0", "0", "0"),
  momid = c("4", "5", "7", "9", "0", "0", "0", "0", "0", "0"),
  rel_df = data.frame(
      id1 = "10",
      id2 = "3",
      code = "Spouse"
  )
)

data(sampleped)
ped1 <- Pedigree(sampleped[sampleped$famid == "1",])
ped1 <- num_child(ped1, reset = TRUE)
summary(ped(ped1))

Description

Convert a Pedigree to a legend data frame for it to be plotted afterwards with plot_fromdf().

Usage

## S4 method for signature 'Pedigree'
ped_to_legdf(obj, boxh = 1, boxw = 1, cex = 1, adjx = 0, adjy = 0)

Arguments

Details

The data frame contains the following columns:

• x0, y0, x1, y1: coordinates of the elements

• type: type of the elements

• fill: fill color of the elements

• border: border color of the elements

• angle: angle of the shading of the elements

• density: density of the shading of the elements

• cex: size of the elements

• label: label of the elements

• tips: tips of the elements (used for the tooltips)

• adjx: horizontal text adjustment of the labels

• adjy: vertical text adjustment of the labels

All those columns are used by plot_fromdf() to plot the graph.

Value

A list containing the legend data frame and the user coordinates.

Examples

data("sampleped")
ped <- Pedigree(sampleped)
leg_df <- ped_to_legdf(ped)
summary(leg_df$df)
plot_fromdf(leg_df$df, usr = c(-1,15,0,7))

Description

Convert a Pedigree to a data frame with all the elements and their characteristic for them to be plotted afterwards with plot_fromdf().

Usage

## S4 method for signature 'Pedigree'
ped_to_plotdf(
  obj,
  packed = TRUE,
  width = 6,
  align = c(1.5, 2),
  subreg = NULL,
  cex = 1,
  symbolsize = cex,
  pconnect = 0.5,
  branch = 0.6,
  aff_mark = TRUE,
  label = NULL,
  ...
)

Arguments

Details

The data frame contains the following columns:

• x0, y0, x1, y1: coordinates of the elements

• type: type of the elements

• fill: fill color of the elements

• border: border color of the elements

• angle: angle of the shading of the elements

• density: density of the shading of the elements

• cex: size of the elements

• label: label of the elements

• tips: tips of the elements (used for the tooltips)

• adjx: horizontal text adjustment of the labels

• adjy: vertical text adjustment of the labels

All those columns are used by plot_fromdf() to plot the graph.

Value

A list containing the data frame and the user coordinates.

See Also

plot_fromdf() ped_to_legdf()

Examples

data(sampleped)
ped1 <- Pedigree(sampleped[sampleped$famid == 1,])
plot_df <- ped_to_plotdf(ped1)
summary(plot_df$df)
plot_fromdf(plot_df$df, usr = plot_df$par_usr$usr,
    boxh = plot_df$par_usr$boxh, boxw = plot_df$par_usr$boxw
)

Description

S4 class to represent the identity informations of the individuals in a pedigree.

Constructor :

You either need to provide a vector of the same size for each slot or a data.frame with the corresponding columns.

The metadata will correspond to the columns that do not correspond to the Ped slots.

Usage

## S4 method for signature 'data.frame'
Ped(obj, cols_used_init = FALSE, cols_used_del = FALSE)

## S4 method for signature 'character_OR_integer'
Ped(
  obj,
  sex,
  dadid,
  momid,
  famid = NA,
  steril = NA,
  status = NA,
  avail = NA,
  affected = NA,
  missid = NA_character_
)

Arguments

Details

The minimal needed informations are id, dadid, momid and sex. The other slots are used to store recognized informations. Additional columns can be added to the Ped object and will be stored in the elementMetadata slot of the Ped object.

Value

A Ped object.

Slots

id

A character vector with the id of the individuals.

dadid

A character vector with the id of the father of the individuals.

momid

A character vector with the id of the mother of the individuals.

sex

An ordered factor vector for the sex of the individuals (i.e. male < female < unknown < terminated).

famid

A character vector with the family identifiers of the individuals (optional).

steril

A logical vector with the sterilisation status of the individuals (i.e. FALSE = not sterilised, TRUE = sterilised, NA = unknown).

status

A logical vector with the affection status of the individuals (i.e. FALSE = alive, TRUE = dead, NA = unknown).

avail

A logical vector with the availability status of the individuals (i.e. FALSE = not available, TRUE = available, NA = unknown).

affected

A logical vector with the affection status of the individuals (i.e. FALSE = not affected, TRUE = affected, NA = unknown).

useful

A logical vector with the usefulness status of the individuals (i.e. FALSE = not useful, TRUE = useful).

isinf

A logical vector indicating if the individual is informative or not (i.e. FALSE = not informative, TRUE = informative).

kin

A numeric vector with minimal kinship value between the individuals and the useful individuals.

num_child_tot

A numeric vector with the total number of children of the individuals.

num_child_dir

A numeric vector with the number of children of the individuals.

num_child_ind

A numeric vector with the number of children of the individuals.

elementMetadata

A DataFrame with the additional metadata columns of the Ped object.

metadata

Meta informations about the pedigree.

Accessors

For all the following accessors, the x parameters is a Ped object. Each getters return a vector of the same length as x with the values of the corresponding slot. For each getter, you have a setter with the same name, to be use as slot(x) <- value. The value parameter is a vector of the same length as x, except for the mcols() accessors where value is a list or a data.frame with each elements with the same length as x.

• id(x) : Individuals identifiers

• dadid(x) : Individuals' father identifiers

• momid(x) : Individuals' mother identifiers

• famid(x) : Individuals' family identifiers

• sex(x) : Individuals' gender

• affected(x) : Individuals' affection status

• avail(x) : Individuals' availability status

• status(x) : Individuals' death status

• isinf(x) : Individuals' informativeness status

• kin(x) : Individuals' kinship distance to the informative individuals

• useful(x) : Individuals' usefullness status

• mcols(x) : Individuals' metadata

Generics

• summary(x): Compute the summary of a Ped object

• show(x): Convert the Ped object to a data.frame and print it with its summary.

• as.list(x): Convert a Ped object to a list with the metadata columns at the end.

• as.data.frame(x): Convert a Ped object to a data.frame with the metadata columns at the end.

• subset(x, i, del_parents = FALSE, keep = TRUE): Subset a Ped object based on the individuals identifiers given.

  • i : A vector of individuals identifiers to keep.

  • del_parents : A logical value indicating if the parents of the individuals should be deleted.

  • keep : A logical value indicating if the individuals should be kept or deleted.

See Also

Pedigree()

Examples

data(sampleped)
Ped(sampleped)

Ped(
    obj = c("1", "2", "3", "4", "5", "6"),
    dadid = c("4", "4", "6", "0", "0", "0"),
    momid = c("5", "5", "5", "0", "0", "0"),
    sex = c(1, 2, 3, 1, 2, 1),
    missid = "0"
)

Description

A pedigree is a ensemble of individuals linked to each other into a family tree. A Pedigree object store the informations of the individuals and the special relationships between them. It also permit to store the informations needed to plot the pedigree (i.e. scales and hints).

Constructor :

Main constructor of the package. This constructor help to create a Pedigree object from different data.frame or a set of vectors.

If any errors are found in the data, the function will return the data.frame with the errors of the Ped object and the Rel object.

Usage

Pedigree(obj, ...)

## S4 method for signature 'character_OR_integer'
Pedigree(
  obj,
  dadid,
  momid,
  sex,
  famid = NA,
  avail = NULL,
  affected = NULL,
  status = NULL,
  steril = NULL,
  rel_df = NULL,
  missid = NA_character_,
  col_aff = "affection",
  normalize = TRUE,
  ...
)

## S4 method for signature 'data.frame'
Pedigree(
  obj = data.frame(indId = character(), fatherId = character(), motherId = character(),
    gender = numeric(), family = character(), available = numeric(), vitalStatus =
    numeric(), affection = numeric(), sterilisation = numeric()),
  rel_df = data.frame(id1 = character(), id2 = character(), code = numeric(), famid =
    character()),
  cols_ren_ped = list(indId = "id", fatherId = "dadid", motherId = "momid", family =
    "famid", gender = "sex", sterilisation = "steril", affection = "affected", available
    = "avail", vitalStatus = "status"),
  cols_ren_rel = list(id1 = "indId1", id2 = "indId2", famid = "family"),
  hints = list(horder = NULL, spouse = NULL),
  normalize = TRUE,
  missid = NA_character_,
  col_aff = "affection",
  ...
)

Arguments

Details

If the normalization is set to TRUE, then the data will be standardized using the function norm_ped() and norm_rel().

If a data.frame is given, the columns names needed will depend if the normalization is selected or not. If the normalization is selected, the columns names needed are as follow and if not the columns names needed are in parenthesis:

• indID: the individual identifier (id)

• fatherId: the identifier of the biological father (dadid)

• motherId: the identifier of the biological mother (momid)

• gender: the sex of the individual (sex)

• family: the family identifier of the individual (famid)

• sterilisation: the sterilisation status of the individual (steril)

• available: the availability status of the individual (avail)

• vitalStatus: the death status of the individual (status)

• affection: the affection status of the individual (affected)

• ...: other columns that will be stored in the elementMetadata slot

The minimum columns required are :

• indID / id

• fatherId / dadid

• motherId / momid

• gender / sex

The family / famid column can also be used to specify the family of the individuals and will be merge to the indId / id field separated by an underscore. The columns sterilisation, available, vitalStatus, affection will be transformed with the vect_to_binary() function when the normalisation is selected. If you do not use the normalisation, the columns will be checked to be 0 or 1.

If affected is a data.frame, col_aff will be overwritten by the column names of the data.frame.

Value

A Pedigree object.

Slots

ped

A Ped object for the identity informations. See Ped() for more informations.

rel

A Rel object for the special relationships. See Rel() for more informations.

scales

A Scales object for the filling and bordering colors used in the plot. See Scales() for more informations.

hints

A Hints object for the ordering of the individuals in the plot. See Hints() for more informations.

Accessors

For all the following accessors, the x parameters is a Pedigree object. Each getters return a vector of the same length as x with the values of the corresponding slot.

• famid(x) : Get the family identifiers of a Pedigree object. This function is a wrapper around famid(ped(x)).

• ped(x, slot) : Get the value of a specific slot of the Ped object

• ped(x) : Get the Ped object

• ped(x, slot) <- value : Set the value of a specific slot of the Ped object Wrapper of slot(ped(x)) <- value

• ped(x) <- value : Set the Ped object

• mcols(x) : Get the metadata of a Pedigree object. This function is a wrapper around mcols(ped(x)).

• mcols(x) <- value : Set the metadata of a Pedigree object. This function is a wrapper around mcols(ped(x)) <- value.

• rel(x, slot) : Get the value of a specific slot of the Rel object

• rel(x) : Get the Rel object

• rel(x, slot) <- value : Set the value of a specific slot of the Rel object Wrapper of slot(rel(x)) <- value

• rel(x) <- value : Set the Rel object

• scales(x) : Get the Scales object

• scales(x) <- value : Set the Scales object

• fill(x) : Get the fill data.frame from the Scales object. Wrapper of fill(scales(x))

• fill(x) <- value : Set the fill data.frame from the Scales object. Wrapper of fill(scales(x)) <- value

• border(x) : Get the border data.frame from the Scales object. Wrapper of border(scales(x))

• border(x) <- value : Set the border data.frame from the Scales object. Wrapper of border(scales(x)) <- value

• hints(x) : Get the Hints object

• hints(x) <- value : Set the Hints object

• horder(x) : Get the horder vector from the Hints object. Wrapper of horder(hints(x))

• horder(x) <- value : Set the horder vector from the Hints object. Wrapper of horder(hints(x)) <- value

• spouse(x) : Get the spouse data.frame from the Hints object. Wrapper of spouse(hints(x)).

• spouse(x) <- value : Set the spouse data.frame from the Hints object. Wrapper of spouse(hints(x)) <- value.

Generics

• length(x): Get the length of a Pedigree object. Wrapper of length(ped(x)).

• show(x): Print the information of the Ped and Rel object inside the Pedigree object.

• summary(x): Compute the summary of the Ped and Rel object inside the Pedigree object.

• as.list(x): Convert a Pedigree object to a list

• subset(x, i, keep = TRUE): Subset a Pedigree object based on the individuals identifiers given.

  • i : A vector of individuals identifiers to keep.

  • del_parents : A logical value indicating if the parents of the individuals should be deleted.

  • keep : A logical value indicating if the individuals should be kept or deleted.

• x[i, del_parents, keep]: Subset a Pedigree object based on the individuals identifiers given.

See Also

Pedigree(), Ped(), Rel(), Scales(), Hints()

Ped(), Rel(), Scales()

Examples

Pedigree(
   obj = c("1", "2", "3", "4", "5", "6"),
   dadid = c("4", "4", "6", "0", "0", "0"),
   momid = c("5", "5", "5", "0", "0", "0"),
   sex = c(1, 2, 3, 1, 2, 1),
   avail = c(0, 1, 0, 1, 0, 1),
   affected = matrix(c(
       0, 1, 0, 1, 0, 1,
       1, 1, 1, 1, 1, 1
   ), ncol = 2),
   col_aff = c("aff1", "aff2"),
   missid = "0",
   rel_df = matrix(c(
      "1", "2", 2
   ), ncol = 3, byrow = TRUE),
)

data(sampleped)
Pedigree(sampleped)

Description

Given a vector of length n, generate all possible permutations of the numbers 1 to n. This is a recursive routine, and is not very efficient.

Usage

permute(x)

Arguments

Value

A matrix with n cols and n! rows

Description

This function is used to create a plot from a data.frame.

If ggplot_gen = TRUE, the plot will be generated with ggplot2 and will be returned invisibly.

Usage

plot_fromdf(
  df,
  usr = NULL,
  title = NULL,
  ggplot_gen = FALSE,
  boxw = 1,
  boxh = 1,
  add_to_existing = FALSE
)

Arguments

Value

an invisible ggplot object and a plot on the current plotting device

Examples

data(sampleped)
ped1 <- Pedigree(sampleped[sampleped$famid == 1,])
lst <- ped_to_plotdf(ped1)
#plot_fromdf(lst$df, lst$par_usr$usr,
#     boxw = lst$par_usr$boxw, boxh = lst$par_usr$boxh
#)

Description

This function is used to plot a Pedigree object.

It is a wrapper for plot_fromdf() and ped_to_plotdf() as well as ped_to_legdf() if legend = TRUE.

Usage

## S4 method for signature 'Pedigree,missing'
plot(
  x,
  aff_mark = TRUE,
  label = NULL,
  ggplot_gen = FALSE,
  cex = 1,
  symbolsize = 1,
  branch = 0.6,
  packed = TRUE,
  align = c(1.5, 2),
  width = 6,
  title = NULL,
  subreg = NULL,
  pconnect = 0.5,
  fam_to_plot = 1,
  legend = FALSE,
  leg_cex = 0.8,
  leg_symbolsize = 0.5,
  leg_loc = NULL,
  leg_adjx = 0,
  leg_adjy = 0,
  ...
)

Arguments

Details

Two important parameters control the looks of the result. One is the user specified maximum width. The smallest possible width is the maximum number of subjects on a line, if the user's suggestion is too low it is increased to 1 + that amount (to give just a little wiggle room).

To make a Pedigree where all children are centered under parents simply make the width large enough, however, the symbols may get very small.

The second is align, a vector of 2 alignment parameters a and b. For each set of siblings at a set of locations x and with parents at p=c(p1,p2) the alignment penalty is

$(1/k^a)\sum{i=1}{k} [(x_i - (p1+p2)/2)]^2$

$\sum(x- \overline(p))^2/(k^a)$

Where k is the number of siblings in the set.

When a = 1 moving a sibship with k sibs one unit to the left or right of optimal will incur the same cost as moving one with only 1 or two sibs out of place.

If a = 0 then large sibships are harder to move than small ones, with the default value a = 1.5 they are slightly easier to move than small ones. The rationale for the default is as long as the parents are somewhere between the first and last siblings the result looks fairly good, so we are more flexible with the spacing of a large family. By tethering all the sibs to a single spot they are kept close to each other. The alignment penalty for spouses is $b(x_1 - x_2)^2$, which tends to keep them together. The size of b controls the relative importance of sib-parent and spouse-spouse closeness.

Value

an invisible list containing

• df : the data.frame used to plot the Pedigree

• par_usr : the user coordinates used to plot the Pedigree

• ggplot : the ggplot object if ggplot_gen = TRUE

Side Effects

Creates plot on current plotting device.

See Also

Pedigree()

Examples

data(sampleped)
pedAll <- Pedigree(sampleped)
#plot(pedAll)

Description

Create a list of x and y coordinates for a polygon with a given number of slices and a list of coordinates for the polygon.

Usage

polyfun(nslice, coor)

Arguments

Value

a list of x and y coordinates

Examples

polyfun(2, list(
     x = c(-0.5, -0.5, 0.5, 0.5),
     y = c(-0.5, 0.5, 0.5, -0.5),
     theta = -c(3, 5, 7, 9) * pi / 4
))

Description

Create a list of polygonal elements with x, y coordinates and theta for the square, circle, diamond and triangle. The number of slices in each element can be specified.

Usage

polygons(nslice = 1)

Arguments

Value

a list of polygonal elements with x, y coordinates and theta by slice.

Examples

polygons()
polygons(4)

Description

Relationship code variable to ordered factor

Usage

rel_code_to_factor(code)

Arguments

Value

an ordered factor vector containing the transformed variable "MZ twin" < "DZ twin" < "UZ twin" < "Spouse"

Examples

rel_code_to_factor(c(1, 2, 3, 4, "MZ twin", "DZ twin", "UZ twin", "Spouse"))

Description

S4 class to represent the special relationships in a Pedigree.

Constructor :

You either need to provide a vector of the same size for each slot or a data.frame with the corresponding columns.

Usage

## S4 method for signature 'data.frame'
Rel(obj)

## S4 method for signature 'character_OR_integer'
Rel(obj, id2, code, famid = NA_character_)

Arguments

Details

A Rel object is a list of special relationships between individuals in the pedigree. It is used to create a Pedigree object. The minimal needed informations are id1, id2 and code.

If a famid is provided, the individuals id will be aggregated to the famid character to ensure the uniqueness of the id.

Value

A Rel object.

Slots

id1

A character vector with the id of the first individual.

id2

A character vector with the id of the second individual.

code

An ordered factor vector with the code of the special relationship.

(i.e. ⁠MZ twin⁠ < ⁠DZ twin⁠ < ⁠UZ twin⁠ < Spouse).

famid

A character vector with the famid of the individuals.

Accessors

For all the following accessors, the x parameters is a Rel object. Each getters return a vector of the same length as x with the values of the corresponding slot.

• code(x) : Relationships' code

• id1(x) : Relationships' first individuals' identifier

• id2(x) : Relationships' second individuals' identifier

• famid(x) : Relationships' individuals' family identifier

• famid(x) <- value : Set the relationships' individuals' family identifier

  • value : A character or integer vector of the same length as x with the family identifiers

Generics

• summary(x): Compute the summary of a Rel object

• show(x): Convert the Rel object to a data.frame and print it with its summary.

• as.list(x): Convert a Rel object to a list

• as.data.frame(x): Convert a Rel object to a data.frame

• subset(x, i, keep = TRUE): Subset a Rel object based on the individuals identifiers given.

  • i : A vector of individuals identifiers to keep.

  • keep : A logical value indicating if the individuals should be kept or deleted.

See Also

Pedigree()

Examples

rel_df <- data.frame(
   id1 = c("1", "2", "3"),
   id2 = c("2", "3", "4"),
   code = c(1, 2, 3)
)
Rel(rel_df)

Rel(
   obj = c("1", "2", "3"),
   id2 = c("2", "3", "4"),
   code = c(1, 2, 3)
)

Description

Small sample pedigree data set for testing purposes.

Usage

data("sampleped")

Format

A data frame with 55 observations, one line per subject, on the following 7 variables.

• famid : Family identifier

• id : Subject identifier

• dadid : Identifier of the father, if the father is part of the data set; zero otherwise

• momid : Identifier of the mother, if the mother is part of the data set; zero otherwise

• sex : 1 for male or 2 for female

• affected : 1 or 0

• avail : 1 or 0

Details

This is a small fictive pedigree data set, with 55 individuals in 2 families. The aim was to create a data set with a variety of pedigree structures.

Examples

data("sampleped")
ped <- Pedigree(sampleped)
summary(ped)
#plot(ped)

Description

A Scales object is a list of two data.frame. The first one is used to represent the affection status of the individuals and therefore the filling of the individuals in the pedigree plot. The second one is used to represent the availability status of the individuals and therefore the border color of the individuals in the pedigree plot.

Constructor :

You need to provide both fill and border in the dedicated parameters. However this is usually done using the generate_colors() function with a Pedigree object.

Usage

Scales(fill, border)

## S4 method for signature 'data.frame,data.frame'
Scales(fill, border)

Arguments

Value

A Scales object.

Slots

fill

A data.frame with the informations for the affection status. The columns needed are:

• 'order': the order of the affection to be used

• 'column_values': name of the column containing the raw values in the Ped object

• 'column_mods': name of the column containing the mods of the transformed values in the Ped object

• 'mods': all the different mods

• 'labels': the corresponding labels of each mods

• 'affected': a logical value indicating if the mod correspond to an affected individuals

• 'fill': the color to use for this mods

• 'density': the density of the shading

• 'angle': the angle of the shading

border

A data.frame with the informations for the availability status. The columns needed are:

• 'column_values': name of the column containing the raw values in the Ped object

• 'column_mods': name of the column containing the mods of the transformed values in the Ped object

• 'mods': all the different mods

• 'labels': the corresponding labels of each mods

• 'border': the color to use for this mods

Accessors

• fill(x) : Get the fill data.frame

• fill(x) <- value : Set the fill data.frame

• border(x) : Get the border data.frame

• border(x) <- value : Set the border data.frame

Generics

• as.list(x): Convert a Scales object to a list

See Also

Pedigree()

generate_colors()

Examples

Scales(
   fill = data.frame(
      order = 1,
      column_values = "affected",
      column_mods = "affected_mods",
      mods = c(0, 1),
      labels = c("unaffected", "affected"),
      affected = c(FALSE, TRUE),
      fill = c("white", "red"),
      density = c(NA, 20),
      angle = c(NA, 45)
   ),
   border = data.frame(
      column_values = "avail",
      column_mods = "avail_mods",
      mods = c(0, 1),
      labels = c("not available", "available"),
      border = c("black", "blue")
   )
)