Package 'catmap'

catmap: Case-Control and TDT Meta-Analysis Package

Description

This package conducts fixed-effects (with inverse variance weighting) and random-effects [DerSimonian and Laird (1986)] meta-analyses of case-control or family-based (TDT) genetic data. In addition, catmap performs meta-analyses which combine these two types of study designs. Specifically, this package implements a fixed-effects model [Kazeem and Farrall (2005)] and a random-effects model [Nicodemus (2008)] for combined studies.

Usage

catmap(dataset, ci = 0.95, printout = FALSE)

Arguments

dataset	A data.frame, matrix, or file location. The input should have the following column names: name, study, t, nt, caserisk, controlrisk, casenotrisk, controlnotrisk. For a file location, provide the data in a tab-delimited format. Note that the header must have these exact columns and all cells in the table must have an entry, even if the entry is 0 or NA. See catmapdata for an example.
ci	A numeric value. The confidence level for confidence intervals; 0 < ci < 1.
printout	A boolean. Toggles whether a text file of the models and Q statistic results should get saved to the working directory.

Details

Use the output of the catmap function to generate figures using secondary functions. These secondary functions produce output as either a txt file, a pdf plot, or both.

A standard forest plot is available via catmap.forest. A funnel plot is available via catmap.funnel. However, no formal test of publication bias is available (see [Ioannidis and Trikalinos (2007)]).

In addition, cumulative meta-analyses over time (catmap.cumulative) and leave-one-out sensitivity analyses (catmap.sense) are available for the fixed-effects estimates and random-effects estimates.

Author(s)

Algorithm designed and implemented by Kristin K. Nicodemus. Code modified and updated by Thom Quinn.

See Also

catmap, catmap.forest, catmap.sense, catmap.cumulative, catmap.funnel

Examples

data(catmapdata)
catmapobject <- catmap(catmapdata, 0.95, TRUE)

---

catmap: Cumulative Meta-Analysis

Description

The catmap.cumulative conducts cumulative meta-analyses and creates plots of Odds Ratios (OR) and Confidence Intervals (CI) using a fixed-effects or random-effects model. Note that studies should be listed in chronological order in the input file! This function does not re-order studies by publication year! Also note that random-effects estimates are not defined for a single (i.e., the first) study.

Usage

catmap.cumulative(catmapobject, fe.forest = FALSE, re.forest = FALSE,
  printout = FALSE)

Arguments

catmapobject	A catmap object created by catmap.
fe.forest	A boolean. Toggles whether the forest plot should get saved to the current working directory.
re.forest	A boolean. Toggles whether the forest plot should get saved to the current working directory.
printout	A boolean. Toggles whether a text file of the models and Q statistic results should get saved to the working directory.

Author(s)

Algorithm designed and implemented by Kristin K. Nicodemus. Code modified and updated by Thom Quinn.

See Also

catmap, catmap.forest, catmap.sense, catmap.cumulative, catmap.funnel

Examples

data(catmapdata)
catmapobject <- catmap(catmapdata, 0.95, TRUE)
catmap.cumulative(catmapobject, FALSE, FALSE, FALSE)

---

catmap: Forest Plot

Description

The catmap.forest creates forest plots of the individual study Odds Ratios (OR) and Confidence Intervals (CI). It then summarizes the data using a fixed-effects or random-effects pooled OR and CI.

Usage

catmap.forest(catmapobject, fe.forest = FALSE, re.forest = FALSE)

Arguments

catmapobject	A catmap object created by catmap.
fe.forest	A boolean. Toggles whether the forest plot should get saved to the current working directory.
re.forest	A boolean. Toggles whether the forest plot should get saved to the current working directory.

Author(s)

Algorithm designed and implemented by Kristin K. Nicodemus. Code modified and updated by Thom Quinn.

See Also

catmap, catmap.forest, catmap.sense, catmap.cumulative, catmap.funnel

Examples

data(catmapdata)
catmapobject <- catmap(catmapdata, 0.95, TRUE)
catmap.forest(catmapobject, TRUE, TRUE)

---

catmap: Funnel Plot

Description

The catmap.funnel creates a funnel plot of the individual Log Odds Ratio against the standard error of the Log Odds Ratio. The vertical line indicates the combined Log Odds Ratio. Per the metafor package, "A pseudo confidence interval region is drawn around this value with bounds equal to +/- 1.96 SE".

Usage

catmap.funnel(catmapobject, funnel = FALSE)

Arguments

catmapobject	A catmap object created by catmap.
funnel	A boolean. Toggles whether the funnel plot should get saved to the current working directory.

Author(s)

Algorithm designed and implemented by Kristin K. Nicodemus. Code modified and updated by Thom Quinn.

See Also

catmap, catmap.forest, catmap.sense, catmap.cumulative, catmap.funnel

Examples

data(catmapdata)
catmapobject <- catmap(catmapdata, 0.95, TRUE)
catmap.funnel(catmapobject, TRUE)

---

catmap: Leave-One-Out Sensitivity Analysis

Description

The catmap.sense conducts leave-one-out sensitivity analyses and creates plots of Odds Ratios (OR) and Confidence Intervals (CI) using a fixed-effects or random-effects model.

Usage

catmap.sense(catmapobject, fe.forest = FALSE, re.forest = FALSE,
  printout = FALSE)

Arguments

catmapobject	A catmap object created by catmap.
fe.forest	A boolean. Toggles whether the forest plot should get saved to the current working directory.
re.forest	A boolean. Toggles whether the forest plot should get saved to the current working directory.
printout	A boolean. Toggles whether a text file of the models and Q statistic results should get saved to the working directory.

Author(s)

Algorithm designed and implemented by Kristin K. Nicodemus. Code modified and updated by Thom Quinn.

See Also

catmap, catmap.forest, catmap.sense, catmap.cumulative, catmap.funnel

Examples

data(catmapdata)
catmapobject <- catmap(catmapdata, 0.95, TRUE)
catmap.sense(catmapobject, FALSE, FALSE, FALSE)

---

Example catmap Data

Description

An example data set for use with catmap. All input data should have the header as part of the file and either 0 or NA values for entries not relevant to that particular study design. For example, TDT studies should have the caserisk, controlrisk, casenotrisk and controlnotrisk values set to either 0 or NA.

Usage

catmapdata

Format

A data.frame with 5 observations and 8 variables.

• name: a factor with study name and optionally year of publication. NOTE: if year of publication is included there must be no space between study name and year. A comma or underscore works nicely (e.g., Abrams,2001 Peter,2002 Todd,2003 Wei,2007 Yu,2007)

• study: a numeric vector containing 1 if the study is TDT and 2 if the study is case-control

• t: a numeric vector containing counts of alleles transmitted in the TDT study

• nt: a numeric vector containing counts of alleles not transmitted in the TDT study

• caserisk: a numeric vector containing counts of risk alleles in cases

• controlrisk: a numeric vector containing counts of risk alleles in controls

• casenotrisk: a numeric vector containing counts of non-risk alleles in cases

• controlnotrisk: a numeric vector containing counts of non-risk alleles in controls

Author(s)

Algorithm designed and implemented by Kristin K. Nicodemus. Code modified and updated by Thom Quinn.

See Also

catmap, catmap.forest, catmap.sense, catmap.cumulative, catmap.funnel

---

Make Forest Plot

Description

A back-end wrapper function used to make forest plots.

Usage

makeForest(catmapobject, summary = "", main = "Main Title",
  mean = exp(catmapobject$logOR), lower = catmapobject$lbci.fe,
  upper = catmapobject$ubci.fe, study = c("Study", sub(",", " ",
  catmapobject$studyname)))

Arguments

catmapobject	A catmap object created by catmap.
summary	A character string. The kind of summary statistic to plot. Select from "fixed" or "random".
main	A character string. The figure title.
mean, lower, upper, study	Numeric or character vectors. Used to guide the construction of the forest plot.

---