--- title: Pedigree object author: Terry Therneau, Elizabeth Atkinson, Louis Le Nézet date: '`r format(Sys.time(),"%d %B, %Y")`' output: BiocStyle::html_document: toc: true toc_depth: 2 header-includes: \usepackage{tabularx} vignette: | %\VignetteIndexEntry{Pedigree object} %\VignetteEncoding{UTF-8} %\VignetteEngine{knitr::rmarkdown} --- Introduction =============== The pedigree routines came out of a simple need -- to quickly draw a Pedigree structure on the screen, within R, that was "good enough" to help with debugging the actual routines of interest, which were those for fitting mixed effecs Cox models to large family data. As such the routine had compactness and automation as primary goals; complete annotation (monozygous twins, multiple types of affected status) and most certainly elegance were not on the list. Other software could do that much better. It therefore came as a major surprise when these routines proved useful to others. Through their constant feedback, application to more complex pedigrees, and ongoing requests for one more feature, the routine has become what it is today. This routine is still not suitable for really large pedigrees, nor for heavily inbred ones such as in animal studies, and will likely not evolve in that way. The authors fondest hope is that others will pick up the project. Pedigree Constructor ======================== ## Arguments The Pedigree function is the first step, creating an object of class Pedigree. It accepts the following input - **ped_df** A dataframe containing the columns - $indId$ A numeric or character vector of subject identifiers. - $fatherId$ The identifier of the father. - $motherId$ The identifier of the mother. - $gender$ The gender of the individual. This can be a numeric variable with codes of 1=male, 2=female, 3=unknown, 4=terminated, or NA=unknown. A character or factor variable can also be supplied containing the above; the string may be truncated and of arbitrary case. - $available$ Optional, a numeric variable with 0 = unavailable and 1 = available. - $affected$ Optional, a numeric variable with 0 = unaffected and 1 = affected. - $status$ Optional, a numeric variable with 0 = censored and 1 = dead. - $famid$ Optional, a numeric or character vector of family identifiers. - $steril$ Optional, a numeric variable with 0 = not steril and 1 = steril. - **rel_df** Optional, a data frame with three columns or four columns. - $indId1$ identifier values of the subject pairs - $indId2$ identifier values of the subject pairs - $code$ relationship codification : 1 = Monozygotic twin, 2=Dizygotic twin, 3= twin of unknown zygosity, 4 = Spouse. - $famid$ Optional, a numeric or character vector of family identifiers. - **cols_ren_ped** Optional, a named list for the renaming of the **ped_df** dataframe - **cols_ren_rel** Optional, a named list for the renaming of the **rel_df** dataframe - **normalize** Optional, a logical to know if the data should be normalised. - **hints** Optional, a list containing the horder in which to plot the individuals and the matrix of the spouse. ## Notes Note that a factor variable is not listed as one of the choices for the subject identifier. This is on purpose. Factors were designed to accomodate character strings whose values came from a limited class -- things like race or gender, and are not appropriate for a subject identifier. All of their special properties as compared to a character variable turn out to be backwards for this case, in particular a memory of the original level set when subscripting is done. However, due to the awful decision early on in S to automatically turn every character into a factor --- unless you stood at the door with a club to head the package off --- most users have become ingrained to the idea of using them for every character variable. (I encourage you to set the global option `stringsAsFactors = FALSE` to turn off autoconversion -- it will measurably improve your R experience). Therefore, to avoid unnecessary hassle for our users the code will accept a factor as input for the id variables, but the final structure does not retain it. Gender and relation do become factors. Status follows the pattern of the survival routines and remains an integer. ## Column renaming Based on the dataframe given for **ped_df** and **rel_df** and their corresponding named list, the columns are renamed for them to be used correctly. The renaming is done as follow ```{r, column renaming} rel_df <- data.frame( indId1 = c("110", "204"), indId2 = c("112", "205"), code = c(1, 2), family = c("1", "2") ) cols_ren_rel <- list( id1 = "indId1", id2 = "indId2", famid = "family" ) ## Rename columns rel old_cols <- as.vector(unlist(cols_ren_rel)) new_cols <- names(cols_ren_rel) cols_to_ren <- match(old_cols, names(rel_df)) names(rel_df)[cols_to_ren[!is.na(cols_to_ren)]] <- new_cols[!is.na(cols_to_ren)] print(rel_df) ``` ## Normalisation If the normalisation process is selected `normalize = TRUE`, then both dataframe will be checked by their dedicated normalization function. It will ensure that all modalities are written correctly and set up the right way. If a $famid$ column is present in the dataframe, then it will be aggregated to the id of each individual and separated by an ''_'' to ensure the uniqueness of the individuals identifiers. ```{r, normalisation} library(Pedixplorer) data("sampleped") cols <- c("sex", "id", "avail") summary(sampleped[cols]) ped <- Pedigree(sampleped) summary(as.data.frame(ped(ped))[cols]) ``` ### Errors present after the normalisation process If any error is detected after the normalisation process, then the normalised dataframe is gave back to the user with errors column added describing the encountered problems. ```{r, rel_df errors} rel_wrong <- rel_df rel_wrong$code[2] <- "A" df <- Pedigree(sampleped, rel_wrong) print(df) ``` ## Validation Now that the data for the Pedigree object creation are ready, they are given to a new $Pedigree$ object, trigerring the $validation$ process. This validation step will check up for many errors such as: - All necessary columns are present - No duplicated $id$ - All $momid$ and $dadid$ are present in $id$ - $sex$ column only contain "male", "female", "unknown" or "terminated" values - $steril$, $status$, $available$, $affected$ only contains 0, 1 or NA values - Father are males and Mother are females - Twins have same parents and MZ twins have same sex - Hints object is valid and ids contained is in the Ped object - ... Pedigree Class ======================== After validation an $S4$ object is generated. This new concept make it possible to easily setup methods for this new type of object. The controls of the parameters is also more precise. The $Pedigree$ object contains 4 slots, each of them contains a different $S4$ object containing a specific type of information used for the Pedigree construction. - $ped$ a Ped object for the Pedigree information with at least the following slots: - $id$ the identifiers of the individuals - $dadid$ the identifiers of the fathers - $momid$ the identifiers of the mothers - $sex$ the gender of each individuals - $rel$ a Rel object describing all special relationship beetween individuals that can't be descibed in the $ped$ slot. The minimal slots needed are : - $id1$ the identifiers of the 1st individuals - $id2$ the identifiers of the 2nd individuals - $code$ factor describing the type of relationship ("MZ twin", "DZ twin", "UZ twin", "Spouse") - $scales$ a Scales object with two slots : - $fill$ a dataframe describing which modalities in which columns correspond to an affected individuals. Plotting information such as colour, angle and density are also provided - $border$ a dataframe describing which modalities in which columns to use to plot the border of the plot elements. - $hints$ a Hints object with two slots : - $horder$ numeric vector for the ordering of the individuals plotting - $spouse$ a matrix of the spouses For more information on each object: - `help(Ped)` - `help(Rel)` - `help(Scales)` - `help(Hints)` Pedigree accessors ======================== As the Pedigree object is now an $S4$ class, we have made available a number of accessors. Most of them can be used as a getter or as a setter to modify a value in the correponding slot of the object ## For the Pedigree object - Get/Set slots : ped(), rel(), scales(), hints() - Wrapper to the Ped object: famid(), mcols() - Wrapper of the Scales object: fill(), border() - Wrapper of the Hints object: horder(), spouse() ## For the Ped object - Given in input: id(), dadid(), momid(), famid(), sex() - Other infos used : affected(), avail(), status() - Computed : isinf(), kin(), useful() - Metadata : mcols() ## For the Rel object - id1(), id2(), code(), famid() ## For the Scales object - fill(), border() ## For the Hints object - horder(), spouse() ## Focus on mcols() The mcols() accessors is the one you should use to add more informations to your individuals. ```{r, mcols} ped <- Pedigree(sampleped) mcols(ped)[8:12] ## Add new columns as a threshold if identifiers of individuals superior ## to a given threshold for example mcols(ped)$idth <- ifelse(as.numeric(mcols(ped)$indId) < 200, "A", "B") mcols(ped)$idth ``` Pedigree methods ======================== With this new S4 object comes multiple methods to ease the use of it: - plot() - summary() - print() - show() - as.list() - `[` - shrink() - generate_colors() - is_informative() - kindepth() - kinship() - make_famid() - upd_famid_id() - num_child() - unrelated() - useful_inds() ```{r, pedigree methods} ## We can change the family name based on an other column ped <- upd_famid_id(ped, mcols(ped)$idth) ## We can substract a given family pedA <- ped[famid(ped) == "A"] ## Plot it plot(pedA, cex = 0.5) ## Do a summary summary(pedA) ## Coerce it to a list as.list(pedA)[[1]][1:3] ## Shrink it to keep only the necessary information lst1_s <- shrink(pedA, max_bits = 10) plot(lst1_s$pedObj, cex = 0.5) ## Compute the kinship individuals matrix kinship(pedA)[1:10, 1:10] ## Get the useful individuals pedA <- useful_inds(pedA, informative = "AvAf") as.data.frame(ped(pedA))["useful"][1:10,] ``` Session information =================== ```{r} sessionInfo() ```