### Install and load pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Initial setup commands to install and load the pcadapt package in an R session.

```R
install.packages("pcadapt")
library(pcadapt)
```

--------------------------------

### Install pcadapt from GitHub

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/index.html

Use this command to install the latest development version of the pcadapt package directly from GitHub.

```r
remotes::install_github("bcm-uga/pcadapt")
```

--------------------------------

### Install and Load qvalue Package

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Provides commands to install the 'qvalue' package from Bioconductor and load it into the R session. Use this for q-value analysis.

```R
source("https://bioconductor.org/biocLite.R")
biocLite("qvalue")
library(qvalue)
```

--------------------------------

### Run pcadapt Shiny App

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/run.pcadapt.html

Launches the Shiny interface for pcadapt. Ensure the package is installed and loaded before running.

```R
run.pcadapt()
```

--------------------------------

### Load Example Pooled Sequencing Data

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Loads a built-in example dataset of pooled sequencing data from the pcadapt package for demonstration or testing purposes.

```R
path_to_file <- system.file("extdata", "pool3pops", package = "pcadapt")
pool.data <- read.pcadapt(path_to_file, type = "pool")
```

--------------------------------

### Install pcadapt from CRAN

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/index.html

Use this command to install the pcadapt package from the Comprehensive R Archive Network (CRAN).

```r
install.packages("pcadapt")
```

--------------------------------

### GET get.pc

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get.pc.html

Retrieves a data frame mapping genetic markers to their most correlated principal component.

```APIDOC
## get.pc

### Description
Returns a data frame where each row contains the index of the genetic marker and the principal component most correlated with it.

### Parameters
#### Arguments
- **x** (pcadapt object) - Required - An object of class 'pcadapt'.
- **list** (integer vector) - Required - A list of integers corresponding to the indices of the markers of interest.

### Response
- **data.frame** - A data frame containing the marker index and the index of the most associated principal component.
```

--------------------------------

### Get Statistics

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get_statistics.html

Computes chi-squared distributed statistics using provided z-scores and a specified method.

```APIDOC
## GET /statistics/get_statistics

### Description
Returns chi-squared distributed statistics.

### Method
GET

### Endpoint
/statistics/get_statistics

### Parameters
#### Query Parameters
- **zscores** (numeric matrix) - Required - A numeric matrix containing the z-scores.
- **method** (character string) - Required - Specifies the method to be used to compute the p-values. Two statistics are currently available, "mahalanobis", and "componentwise".
- **pass** (boolean vector) - Required - A boolean vector.

### Response
#### Success Response (200)
- **statistics** (list) - A list containing the test statistics and the associated p-values.
```

--------------------------------

### Get Score Color Function Signature

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get.score.color.html

This is the function signature for get.score.color. It takes a single argument 'pop' which specifies the population each individual belongs to.

```R
get.score.color(pop)
```

--------------------------------

### run.pcadapt()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

Launches the Shiny application for pcadapt.

```APIDOC
## run.pcadapt()

### Description
Shiny app.

### Method
Function call
```

--------------------------------

### Launch pcadapt Shiny App

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Launches an interactive Shiny web application for pcadapt analysis, providing a graphical user interface for data loading, parameter selection, and result exploration.

```R
library(pcadapt)

# Launch the Shiny app
run.pcadapt()

# The app opens in your default web browser
# Features include:
# - File upload for various genotype formats
# - Interactive parameter selection (K, min.maf, etc.)
# - Real-time scree plots and score plots
# - Manhattan plots with zoom capability
# - Export results to CSV

# Alternative: calling pcadapt() without arguments also launches the app
# pcadapt()
```

--------------------------------

### Compute pcadapt with K Principal Components

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Initialize the pcadapt analysis by specifying the number of principal components (K) to use. This step is crucial before computing test statistics.

```R
x <- pcadapt(filename, K = 2)
```

--------------------------------

### Write PLINK Bed Files with writeBed

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Demonstrates how to write PLINK bed, bim, and fam files from pcadapt or lfmm format files. This function is typically called internally by read.pcadapt but can be used directly.

```R
library(pcadapt)

# writeBed is typically called internally by read.pcadapt
# When you call read.pcadapt on a pcadapt or lfmm file,
# it automatically creates a corresponding .bed file

pcadapt_file <- system.file("extdata", "geno3pops.pcadapt", package = "pcadapt")

# This creates geno3pops.pcadapt.bed, .bim, and .fam files
input <- read.pcadapt(pcadapt_file, type = "pcadapt")

# The bed file is created in the same directory as the input
# bed_path: /path/to/geno3pops.pcadapt.bed

# Direct usage of writeBed (internal function, typically not called directly)
# bed_output <- writeBed(pcadapt_file, is.pcadapt = TRUE)
# For lfmm files:
# bed_output <- writeBed(lfmm_file, is.pcadapt = FALSE)
```

--------------------------------

### read.pcadapt()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

File converter for pcadapt input files.

```APIDOC
## read.pcadapt()

### Description
File Converter.

### Method
Function call
```

--------------------------------

### Get Outlier SNP Indices and Filter by Effect Size

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Retrieve the indices of outlier SNPs identified by a chosen method (e.g., BH) and optionally filter them further based on their effect size.

```r
# Get SNP indices of outliers
outlier_snps <- outliers_bh
cat("Outlier SNP indices:", head(outlier_snps), "...\n")

# Filter outliers by effect size (optional)
strong_outliers <- outliers_bh[abs(x$stat[outliers_bh]) > quantile(x$stat, 0.99, na.rm = TRUE)]
```

--------------------------------

### Load and Run pcadapt Analysis

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Load genetic data from a BED file and perform a PCA-based analysis. Ensure the 'pcadapt' library is loaded.

```r
library(pcadapt)

path_to_file <- system.file("extdata", "geno3pops.bed", package = "pcadapt")
filename <- read.pcadapt(path_to_file, type = "bed")
x <- pcadapt(filename, K = 2)
```

--------------------------------

### Run pcadapt with Direct Frequency Matrix Input

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Provide allele frequency data directly as an R matrix to pcadapt. The matrix should have populations as rows and SNPs as columns, with values representing allele frequencies. This is an alternative to reading from a file.

```r
# For Pool-seq, you can also provide an R matrix directly
# Rows = populations, Columns = SNPs, Values = allele frequencies
freq_matrix <- matrix(runif(300), nrow = 3, ncol = 100)
pool_input <- read.pcadapt(freq_matrix, type = "pool")
res <- pcadapt(pool_input, K = 2)
```

--------------------------------

### Load and Analyze Pool-Seq Data with pcadapt

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Demonstrates how to load pool-sequencing data and perform a principal component analysis using pcadapt. This is useful for identifying genetic differentiation in pooled samples.

```R
pool_file <- system.file("extdata", "pool3pops", package = "pcadapt")
pool_data <- read.pcadapt(pool_file, type = "pool")
res_pool <- pcadapt(pool_data, K = 2)
pool_pc <- get.pc(res_pool, 1:150)
print(pool_pc$PC)
```

--------------------------------

### plot.pcadapt()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

Visualization tool for pcadapt results.

```APIDOC
## plot(pcadapt)

### Description
pcadapt visualization tool.

### Method
Function call
```

--------------------------------

### Load and Run pcadapt on Pool-seq Data

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Load Pool-seq data using read.pcadapt and run the pcadapt analysis. K defaults to n_populations - 1. Ensure the path to the data file is correct.

```r
library(pcadapt)

# Load Pool-seq example data
path_to_file <- system.file("extdata", "pool3pops", package = "pcadapt")
pool_data <- read.pcadapt(path_to_file, type = "pool")

# Run pcadapt on Pool-seq data
# K defaults to n_populations - 1
res_pool <- pcadapt(pool_data, K = 3)
```

--------------------------------

### Perform PCA with pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Run the pcadapt function to perform PCA and compute test statistics. A large number of components (e.g., K=20) is recommended initially to help determine the optimal K.

```R
x <- pcadapt(input = filename, K = 20)
```

--------------------------------

### Run pcadapt on in-memory matrices

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Load genotype data from memory and execute pcadapt analysis.

```R
path_to_file <- system.file("extdata", "SSMPG2017.rds", package = "pcadapt")
genotypes <- readRDS(path_to_file)
print(dim(genotypes))
matrix <- read.pcadapt(genotypes, type = "pcadapt")
res <- pcadapt(matrix, K = 20)
plot(res, option = "screeplot")
```

--------------------------------

### Write PLINK files using writeBed

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/writeBed.html

The writeBed function allows you to convert pcadapt or lfmm files into PLINK format (bed, bim, fam).

```APIDOC
## writeBed

### Description
Function to write bed/bim/fam files from a pcadapt or an lfmm file.

### Method
Not applicable (R function)

### Endpoint
Not applicable (R function)

### Parameters
#### Arguments
- **file** (character) - A pcadapt or lfmm file.
- **is.pcadapt** (boolean) - A boolean value indicating if the input file is in pcadapt format.

### Value
- The input `bedfile` path (character).
```

--------------------------------

### writeBed()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

Writes data to PLINK bed file format.

```APIDOC
## writeBed()

### Description
Write PLINK files.

### Method
Function call
```

--------------------------------

### Accessing pcadapt Results

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Demonstrates how to access various numerical quantities from the output object 'x' of the pcadapt function using the '$' symbol.

```R
summary(x)
```

```R
x$pvalues
```

--------------------------------

### plot.pcadapt Function Documentation

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/plot.pcadapt.html

Documentation for the plot.pcadapt S3 method, used for visualizing pcadapt objects.

```APIDOC
## plot.pcadapt

### Description
This S3 method is designed for objects of class `pcadapt` and provides plotting options for quick visualization. It supports several plot types including screeplot, scores, stat.distribution, manhattan, and qqplot.

### Method
S3 method for `pcadapt` objects

### Endpoint
N/A (This is an R function, not a web API endpoint)

### Parameters
#### Arguments
- **x** (object of class "pcadapt") - The pcadapt object generated with `pcadapt`.
- **...** (...) - Additional arguments.
- **option** (character string) - Specifies the figures to be displayed. Options include: "screeplot", "scores", "stat.distribution", "manhattan", "qqplot". If NULL, all plots are printed. Default is "manhattan".
- **i** (integer) - The principal component onto which individuals are projected when the "scores" option is chosen. Default is 1.
- **j** (integer) - The principal component onto which individuals are projected when the "scores" option is chosen. Default is 2.
- **pop** (list of integers or strings) - Specifies which subpopulation the individuals belong to.
- **col** (list of colors) - Colors to be used in the score plot.
- **chr.info** (list) - Contains the chromosome information for each marker.
- **snp.info** (list) - Contains the names of all genetic markers present in the input.
- **plt.pkg** (character string) - Specifies the package to be used for graphical outputs. Use "plotly" for interactive plots, or "ggplot" for static plots. Default is "ggplot".
- **K** (integer) - The principal component of interest. Only required when using the "componentwise" method.

### Request Example
N/A (This is an R function, not a web API endpoint)

### Response
N/A (This function generates plots, not a structured response)

### Error Handling
N/A
```

--------------------------------

### print_convert

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/print_convert.html

Prints out a summary of the file conversion process for genotype or pooled data.

```APIDOC
## print_convert

### Description
Prints out a summary of the file conversion.

### Parameters
#### Arguments
- **input** (matrix/string) - Required - A genotype matrix or a character string specifying the name of the file to be converted.
- **output** (string) - Required - A character string specifying the name of the output file.
- **M** (integer) - Required - The number of genetic markers present in the data.
- **N** (integer) - Required - The number of individuals present in the data.
- **pool** (integer) - Required - The type of data: 0 for genotype data, 1 for pooled data.
```

--------------------------------

### Visualize pcadapt results

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Generate diagnostic plots for pcadapt objects.

```R
plot(res, option = "screeplot")
```

```R
plot(res, option = "manhattan")
```

--------------------------------

### Access and Visualize Component-wise P-values

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Demonstrates how to access individual component p-values and plot the distribution of test statistics for component-wise analysis. The 'stat.distribution' option visualizes the neutral distribution, and K must be specified.

```R
plot(x_cw, option = "stat.distribution", K = 2)
```

--------------------------------

### Usage of get.pc function

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get.pc.html

Call get.pc with a pcadapt object and a list of marker indices to retrieve correlation data.

```R
get.pc(x, list)
```

--------------------------------

### Convert VCF files using vcf2pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/vcf2pcadapt.html

Use this function to convert VCF files to the pcadapt format. This tool is no longer maintained and users are encouraged to use PLINK instead.

```R
vcf2pcadapt(input, output = "tmp.pcadapt", allele.sep = c("/", "|"))
```

--------------------------------

### read.pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/read.pcadapt.html

Converts genotype matrices or files into formats readable by pcadapt.

```APIDOC
## read.pcadapt

### Description
Converts genotype matrices or files to an appropriate format readable by pcadapt. For a file as input, you can choose to return either a matrix or convert it into bed/bim/fam files.

### Parameters
#### Arguments
- **input** (matrix or character) - Required - A genotype matrix or a character string specifying the name of the file to be converted.
- **type** (character) - Optional - A character string specifying the type of data to be converted from (pcadapt, lfmm, vcf, bed, ped, pool, example).
- **type.out** (character) - Optional - Either a bed file or a standard R matrix.
- **allele.sep** (vector) - Optional - A vector of characters indicating what delimiters are used in VCF files.
```

--------------------------------

### pcadapt()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

Performs Principal Component Analysis for outlier detection.

```APIDOC
## pcadapt()

### Description
Principal Component Analysis for outlier detection.

### Method
Function call
```

--------------------------------

### Analyze Pool-Seq Data with pcadapt

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Read and analyze pool-sequencing data. The K parameter defaults to n_populations - 1 if not specified.

```r
pool_file <- system.file("extdata", "pool3pops", package = "pcadapt")
pool_data <- read.pcadapt(pool_file, type = "pool")
res_pool <- pcadapt(pool_data, K = 3)
```

--------------------------------

### Summarize and Plot pcadapt Results

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

View a summary of the pcadapt analysis results and generate a scree plot to visualize the proportion of variance explained by each principal component. Note that the scree plot may not be informative with a small number of populations.

```r
# View summary
summary(res_pool)

# Scree plot (may not be informative with few populations)
plot(res_pool, option = "screeplot")
```

--------------------------------

### Compute pcadapt with Minimum Minor Allele Frequency Threshold

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Run the pcadapt analysis while setting a minimum minor allele frequency (min.maf) threshold. SNPs with MAF below this threshold will have their p-values reported as NA.

```R
x <- pcadapt(filename, K = 2, min.maf = 0.05)
```

--------------------------------

### Generate Q-Q Plot

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Generates a Q-Q plot to compare the observed p-values against their expected uniform distribution. Requires the 'pcadapt' object 'x'.

```R
plot(x, option = "qqplot")
```

--------------------------------

### Visualize Loadings and Manhattan Plot

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Visualize the distribution of loadings for the first two principal components and display the Manhattan plot to identify potential adaptive signals. This is useful for an initial assessment of the PCA results.

```r
par(mfrow = c(1, 2))
for (i in 1:2) {
  plot(res_final$loadings[, i], pch = 19, cex = 0.3,
       ylab = paste0("Loadings PC", i))
}

plot(res_final, option = "manhattan")
```

--------------------------------

### Define plot.pcadapt method signature

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/plot.pcadapt.html

The S3 method signature for plotting pcadapt objects with various configuration options.

```R
# S3 method for pcadapt
plot(
  x,
  ...,
  option = "manhattan",
  i = 1,
  j = 2,
  pop,
  col,
  chr.info = NULL,
  snp.info = NULL,
  plt.pkg = "ggplot",
  K = NULL
)
```

--------------------------------

### Perform PCA with LD Clumping and Plot Loadings

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Performs PCA with LD clumping and then plots the loadings for the first two principal components. This visualization helps assess the distribution of SNP contributions after LD thinning.

```R
res <- pcadapt(matrix, K = 2, LD.clumping = list(size = 200, thr = 0.1))
par(mfrow = c(1, 2))
for (i in 1:2)
  plot(res$loadings[, i], pch = 19, cex = .3, ylab = paste0("Loadings PC", i))
```

--------------------------------

### Graphical Tools

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Provides functions for visualizing PCA results, including Manhattan plots, Q-Q plots, and histograms of test statistics and p-values.

```APIDOC
## Graphical Tools

### Manhattan Plot
A Manhattan plot displays -log10 of the p-values.

```R
plot(x , option = "manhattan")
```

### Q-Q Plot
The user can also check the expected uniform distribution of the p-values using a Q-Q plot.

```R
plot(x, option = "qqplot")
```

This plot confirms that most of the p-values follow the expected uniform distribution. However, the smallest p-values are smaller than expected confirming the presence of outliers.

### Histograms of the test statistic and of the p-values
An histogram of p-values confirms that most of the p-values follow an uniform distribution. The excess of small p-values indicates the presence of outliers.

```R
hist(x$pvalues, xlab = "p-values", main = NULL, breaks = 50, col = "orange")
```

The presence of outliers is also visible when plotting a histogram of the test statistic DjD_j.

```R
plot(x, option = "stat.distribution")
```
```

--------------------------------

### Convert ped files to pcadapt format

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/ped2pcadapt.html

Use ped2pcadapt to convert input ped files to the pcadapt format. Specify the input and output file names.

```R
ped2pcadapt(input, output)
```

--------------------------------

### Calculate and Identify Outliers using q-values

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Calculates q-values from p-values and identifies SNPs with q-values below a specified alpha threshold, indicating potential outliers. Requires the 'qvalue' package and the 'pcadapt' object 'x'.

```R
library(qvalue)
qval <- qvalue(x$pvalues)$qvalues
alpha <- 0.1
outliers <- which(qval < alpha)
length(outliers)
```

--------------------------------

### bed2matrix()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

Converts a PLINK bed file to a matrix format.

```APIDOC
## bed2matrix()

### Description
Convert a bed to a matrix.

### Method
Function call
```

--------------------------------

### Read genotype data

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Convert genotype files to the required bed format using read.pcadapt. Supported formats include pcadapt, lfmm, vcf, bed, ped, and pool.

```R
path_to_file <- "path_to_directory/foo.lfmm"
filename <- read.pcadapt(path_to_file, type = "lfmm")
```

```R
path_to_file <- system.file("extdata", "geno3pops.bed", package = "pcadapt")
filename <- read.pcadapt(path_to_file, type = "bed")
```

--------------------------------

### pcadapt Function Reference

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/pcadapt.html

Reference for the pcadapt function, including its various methods and parameters for analyzing genetic data.

```APIDOC
## pcadapt Function

### Description
Performs principal component analysis and computes p-values to test for outliers. The test for outliers is based on the correlations between genetic variation and the first K principal components. Handles Pool-seq data.

### Method
S3 methods for different input types: `pcadapt_matrix`, `pcadapt_bed`, `pcadapt_pool`.

### Endpoint
N/A (R function)

### Parameters

#### input
- **input** (object) - The output of function `read.pcadapt`.

#### K
- **K** (integer) - Specifies the number of principal components to retain. Default is 2.

#### method
- **method** (character string) - Specifies the method to be used to compute the p-values. Available options: `"mahalanobis"` (default), `"componentwise"`. Deprecated: `"communality"`.

#### min.maf
- **min.maf** (numeric) - Threshold of minor allele frequencies above which p-values are computed. Default is `0.05`.

#### ploidy
- **ploidy** (integer or NULL) - Number of trials, parameter of the binomial distribution. Default is 2 (diploid). For Pool-seq data, this can be NULL.

#### LD.clumping
- **LD.clumping** (list or NULL) - If not NULL, specifies parameters for SNP thinning. Should be a named list with `$size` (window radius) and `$thr` (squared correlation threshold). Default is `NULL`.

#### pca.only
- **pca.only** (logical) - If TRUE, returns only PCA results before statistic computation. Default is `FALSE`.

#### tol
- **tol** (numeric) - Convergence criterion for `RSpectra::svds()`. Default is `1e-4`.

### Value
- Returns an object of class `pcadapt`.

### Details
Performs PCA on scaled and centered genotype data. Computes test statistics based on the chosen `method`.
- `"mahalanobis"`: Uses robust Mahalanobis distance based on z-scores and their covariance.
- `"componentwise"`: Returns a matrix of z-scores.

For `"mahalanobis"` method, scaled statistics (`chi2_stat`) should follow a chi-squared distribution with K degrees of freedom. For `"componentwise"`, z-scores should follow a chi-squared distribution with 1 degree of freedom.

### Request Example
```R
# Example for matrix input
# data_matrix <- read.pcadapt("your_data.geno", format = "geno", parallel = TRUE)
# result_matrix <- pcadapt(data_matrix, K = 20, method = "mahalanobis")

# Example for bed input
# data_bed <- read.pcadapt("your_data.bed", format = "bed", parallel = TRUE)
# result_bed <- pcadapt(data_bed, K = 20, method = "mahalanobis")

# Example for pool input
# data_pool <- read.pcadapt("your_data.pool", format = "pool", parallel = TRUE)
# result_pool <- pcadapt(data_pool, K = 20, method = "mahalanobis")
```

### Response
#### Success Response (Object of class `pcadapt`)
- **@data** (matrix) - Matrix of statistics.
- **@chi2_stat** (numeric vector) - Chi-squared statistics.
- **@gif** (numeric) - Genomic inflation factor.
- **@pvalues** (numeric vector) - P-values.
- **@loadings** (matrix) - Loadings of the principal components.
- **@eigenvalues** (numeric vector) - Eigenvalues.
- **@pcadapt** (list) - List containing intermediate results.

#### Response Example
```R
# Assuming 'result' is the output object from pcadapt()
# print(result)
# summary(result)
```
```

--------------------------------

### Visualize P-value Distribution

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Create a histogram to visualize the distribution of raw p-values obtained from the pcadapt analysis. This can help in assessing the overall significance.

```r
# Examine p-value distribution
hist(x$pvalues, breaks = 50, col = "orange", xlab = "p-values", main = "P-value Distribution")
```

--------------------------------

### Load and Read Genotype Data for pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Loads genotype data from an RDS file and reads it into a matrix format suitable for pcadapt analysis. Assumes individuals are in columns and SNPs are in rows.

```R
path_to_file <- system.file("extdata", "SSMPG2017.rds", package = "pcadapt")
genotypes <- readRDS(path_to_file)
matrix <- read.pcadapt(genotypes, type = "pcadapt")
```

--------------------------------

### Perform Component-wise Genome Scan

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Reads genotype data and performs a component-wise genome scan using pcadapt. Ensure the path to the genotype file is correct. K specifies the number of principal components to consider.

```R
path_to_file <- system.file("extdata", "geno3pops.bed", package = "pcadapt")
filename <- read.pcadapt(path_to_file, type = "bed")
x_cw <- pcadapt(filename, K = 2, method = "componentwise")
summary(x_cw$pvalues)
```

--------------------------------

### Run PCA-only Mode in pcadapt

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Use pca.only = TRUE to compute PCA results without calculating statistics. This is useful for initial exploration of population structure.

```r
x_pca <- pcadapt(filename, K = 10, pca.only = TRUE)
```

--------------------------------

### Perform PCA and Plot Results

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Performs PCA with a specified number of components (K=4) and plots the results, typically showing the distribution of individuals in the principal component space.

```R
res <- pcadapt(matrix, K = 4)
plot(res)
```

--------------------------------

### Perform Genome Scan after LD Thinning

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Conducts a genome scan using pcadapt after applying SNP thinning due to LD. This analysis aims to identify regions involved in local adaptation.

```R
plot(res)
```

--------------------------------

### Generate a Scree Plot

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Displays the percentage of variance explained by each principal component.

```R
plot(x, option = "screeplot")
```

--------------------------------

### get.pc()

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/index.html

Retrieves the principal component most associated with a genetic marker.

```APIDOC
## get.pc()

### Description
Get the principal component the most associated with a genetic marker.

### Method
Function call
```

--------------------------------

### Component-wise Plots for Specific PCs

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Generate plots for specific principal components when using the 'componentwise' method. This allows for detailed examination of individual component effects.

```r
# Component-wise plots require specifying K
x_cw <- pcadapt(filename, K = 2, method = "componentwise")
plot(x_cw, option = "manhattan", K = 1)  # Manhattan plot for PC1
plot(x_cw, option = "stat.distribution", K = 2)  # Distribution for PC2
plot(x_cw, option = "qqplot", K = 1)
```

--------------------------------

### Perform PCA and outlier detection with pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/pcadapt.html

The primary function for running PCA and computing p-values. It supports different input types and statistical methods for outlier detection.

```R
pcadapt(
  input,
  K = 2,
  method = "mahalanobis",
  min.maf = 0.05,
  ploidy = 2,
  LD.clumping = NULL,
  pca.only = FALSE,
  tol = 1e-04
)
```

```R
# S3 method for pcadapt_matrix
pcadapt(
  input,
  K = 2,
  method = c("mahalanobis", "componentwise"),
  min.maf = 0.05,
  ploidy = 2,
  LD.clumping = NULL,
  pca.only = FALSE,
  tol = 1e-04
)
```

```R
# S3 method for pcadapt_bed
pcadapt(
  input,
  K = 2,
  method = c("mahalanobis", "componentwise"),
  min.maf = 0.05,
  ploidy = 2,
  LD.clumping = NULL,
  pca.only = FALSE,
  tol = 1e-04
)
```

```R
# S3 method for pcadapt_pool
pcadapt(
  input,
  K = (nrow(input) - 1),
  method = "mahalanobis",
  min.maf = 0.05,
  ploidy = NULL,
  LD.clumping = NULL,
  pca.only = FALSE,
  tol
)
```

--------------------------------

### Read Genotype Data with read.pcadapt

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Converts various genotype file formats or in-memory matrices into a format compatible with the pcadapt function.

```r
library(pcadapt)

# Read from a bed file (PLINK binary format)
path_to_file <- system.file("extdata", "geno3pops.bed", package = "pcadapt")
filename <- read.pcadapt(path_to_file, type = "bed")
# Returns: pcadapt_bed object with n=150 individuals, p=1500 SNPs

# Read from a pcadapt format file (individuals in columns, SNPs in rows)
# Returns bed file by default for large datasets
pcadapt_file <- system.file("extdata", "geno3pops.pcadapt", package = "pcadapt")
input <- read.pcadapt(pcadapt_file, type = "pcadapt")

# Read from lfmm format file (individuals in rows, SNPs in columns)
# lfmm_file <- "path/to/data.lfmm"
# input <- read.pcadapt(lfmm_file, type = "lfmm")

# Read genotype matrix directly (individuals in columns, SNPs in rows)
genotypes <- readRDS(system.file("extdata", "SSMPG2017.rds", package = "pcadapt"))
matrix_input <- read.pcadapt(genotypes, type = "pcadapt")
# For matrices with individuals in rows, SNPs in columns, use type = "lfmm"

# Read Pool-seq data (populations x SNPs frequency matrix)
pool_file <- system.file("extdata", "pool3pops", package = "pcadapt")
pool_data <- read.pcadapt(pool_file, type = "pool")
# Returns: pcadapt_pool object ready for population-level analysis

# Convert to in-memory matrix instead of bed file
input_matrix <- read.pcadapt(pcadapt_file, type = "pcadapt", type.out = "matrix")
```

--------------------------------

### pcadapt()

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Performs principal component analysis on genetic data to identify population structure and compute test statistics for outlier detection.

```APIDOC
## pcadapt(input, K, pca.only, method)

### Description
Performs PCA on genetic data. Can be used for PCA-only mode or to compute test statistics for genome scans.

### Parameters
- **input** (object) - Required - The genetic data object (e.g., from read.pcadapt).
- **K** (integer) - Optional - Number of principal components to retain.
- **pca.only** (boolean) - Optional - If TRUE, returns only PCA results without computing statistics.
- **method** (string) - Optional - Method for computing statistics (e.g., "componentwise").
```

--------------------------------

### Generate Score Plot with Population Labels

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Visualize population structure by projecting individuals onto principal components. Use the 'pop' argument to color individuals by their known population.

```R
plot(x, option = "scores", pop = poplist.int)
```

```R
plot(x, option = "scores", pop = poplist.names)
```

--------------------------------

### ped2pcadapt Function

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/ped2pcadapt.html

Converts genetic data files from ped format to pcadapt format.

```APIDOC
## ped2pcadapt

### Description
Converts ped files to the format pcadapt.

### Method
Function Call

### Parameters
#### Arguments
- **input** (character string) - Required - The name of the file to be converted.
- **output** (character string) - Required - The name of the output file.

### Request Example
ped2pcadapt("input_file.ped", "output_file.pcadapt")
```

--------------------------------

### Generate Manhattan Plot

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Generates a Manhattan plot to visualize p-values from the pcadapt analysis. Requires the 'pcadapt' object 'x'.

```R
plot(x , option = "manhattan")
```

--------------------------------

### Generate Q-Q Plot for P-values

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Assess the distribution of observed p-values against expected p-values using a Q-Q plot. This helps identify deviations from the null hypothesis.

```r
# Q-Q plot - observed vs expected p-values
plot(x, option = "qqplot")
```

--------------------------------

### PCAdapt Object Structure

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

The 'x' object returned by the pcadapt function contains various numerical quantities derived from PCA on genotype data. These include scores, singular values, loadings, z-scores, allele frequencies, and statistics for outlier detection.

```APIDOC
## PCAdapt Object Structure

The object `x` returned by the `pcadapt` function contains numerical quantities obtained after performing a PCA on the genotype matrix.

We assume in the following that there are `n` individuals and `L` markers.

*   `scores`: A `(n,K)` matrix corresponding to the projections of the individuals onto each PC.
*   `singular.values`: A vector containing the `K` ordered square root of the proportion of variance explained by each PC.
*   `loadings`: A `(L,K)` matrix containing the correlations between each genetic marker and each PC.
*   `zscores`: A `(L,K)` matrix containing the z-scores.
*   `af`: A vector of size `L` containing allele frequencies of derived alleles where genotypes of `0` are supposed to code for homozygous for the reference allele.
*   `maf`: A vector of size `L` containing minor allele frequencies.
*   `chi2.stat`: A vector of size `L` containing the rescaled statistics `stat/gif` that follow a chi-squared distribution with `K` degrees of freedom.
*   `gif`: A numerical value corresponding to the genomic inflation factor estimated from `stat`.
*   `pvalues`: A vector containing `L` p-values.
*   `pass`: A list of SNPs indices that are kept after exclusion based on the minor allele frequency threshold.
*   `stat`: A vector of size `L` containing squared Mahalanobis distances by default.

All of these elements are accessible using the `$` symbol. For example, the p-values are contained in `x$pvalues`.
```

--------------------------------

### Convert BED to Matrix - pcadapt

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/bed2matrix.html

Converts a BED file to an integer matrix. Ensure the BED file path is correct. The function can infer sample and SNP counts from associated .fam and .bim files if not provided.

```R
bedfile <- system.file("extdata", "geno3pops.bed", package = "pcadapt")
mat <- bed2matrix(bedfile)
dim(mat)
table(mat)
```

--------------------------------

### Histogram of p-values

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Creates a histogram of p-values to visualize their distribution. Useful for identifying an excess of small p-values indicating outliers. Requires the 'pcadapt' object 'x'.

```R
hist(x$pvalues, xlab = "p-values", main = NULL, breaks = 50, col = "orange")
```

--------------------------------

### Outlier Detection Methods

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Methods for identifying outlier SNPs using q-values and the Benjamini-Hochberg procedure.

```APIDOC
## Choosing a Cutoff for Outlier Detection

To provide a list of outliers and choose a cutoff for outlier detection, there are several methods that are listed below from the less conservative one to the more conservative one.

### q-values
The `R` package [qvalue](http://www.bioconductor.org/packages/release/bioc/html/qvalue.html), transforms p-values into q-values. To install and load the package, type the following command lines:

```R
## try http if https is not available
source("https://bioconductor.org/biocLite.R")
biocLite("qvalue")
library(qvalue)
```

For a given \alpha (real valued number between 0 and 1), SNPs with q-values less than \alpha will be considered as outliers with an expected false discovery rate bounded by \alpha. The false discovery rate is defined as the percentage of false discoveries among the list of candidate SNPs. Here is an example of how to provide a list of candidate SNPs for the **geno3pops** data, for an expected false discovery rate lower than `10%`:

```R
library(qvalue)
qval <- qvalue(x$pvalues)$qvalues
alpha <- 0.1
outliers <- which(qval < alpha)
length(outliers)
```

### Benjamini-Hochberg Procedure

```R
padj <- p.adjust(x$pvalues,method="BH")
alpha <- 0.1
outliers <- which(padj < alpha)
length(outliers)
```
```

--------------------------------

### Plot SNP Loadings for PCA Components

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Visualizes the loadings (contributions of each SNP to the principal components) for the first four PCs. Useful for identifying potential regions of Linkage Disequilibrium (LD).

```R
par(mfrow = c(2, 2))
for (i in 1:4)
  plot(res$loadings[, i], pch = 19, cex = .3, ylab = paste0("Loadings PC", i))
```

--------------------------------

### Plot Distribution of Test Statistics

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Visualize the distribution of test statistics and compare it to the expected chi-squared distribution. This is useful for assessing the overall fit of the model.

```r
# Histogram of test statistics with expected chi-squared distribution
plot(x, option = "stat.distribution")
```

--------------------------------

### get.pop.names

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get.pop.names.html

Retrieves population names from a list of population identifiers.

```APIDOC
## get.pop.names

### Description
Retrieves the population names from the population file.

### Method
Function Call

### Parameters
#### Arguments
- **pop** (list of integers or strings) - Required - A list specifying which population the individuals belong to.

### Request Example
get.pop.names(pop = c(1, 1, 2, 2))
```

--------------------------------

### Generate Manhattan Plot for pcadapt Results

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Visualize p-values across the genome using a Manhattan plot. Chromosome information can be optionally provided.

```r
# Manhattan plot (default) - displays -log10(p-values) for each SNP
plot(x, option = "manhattan")

# Manhattan plot with chromosome information
# chr_info <- read.table("snp_positions.txt")$V1  # chromosome numbers
# plot(x, option = "manhattan", chr.info = chr_info)
```

--------------------------------

### Generate Interactive Plots with Plotly

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Create interactive versions of Manhattan and score plots using the plotly package. This allows for dynamic exploration of the results.

```r
# Interactive plots using plotly
plot(x, option = "manhattan", plt.pkg = "plotly")
plot(x, option = "scores", pop = pop_labels, plt.pkg = "plotly")
```

--------------------------------

### Perform PCA and Generate Scree Plot

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Performs Principal Component Analysis (PCA) on genotype data and generates a scree plot to help determine the optimal number of principal components (K).

```R
res <- pcadapt(matrix, K = 20)
plot(res, option = "screeplot")
```

--------------------------------

### Identify Outliers using Benjamini-Hochberg Procedure

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Applies the Benjamini-Hochberg procedure to adjust p-values and identifies SNPs with adjusted p-values below a specified alpha threshold. Requires the 'pcadapt' object 'x'.

```R
padj <- p.adjust(x$pvalues,method="BH")
alpha <- 0.1
outliers <- which(padj < alpha)
length(outliers)
```

--------------------------------

### Convert vcfR genotype matrices to pcadapt format

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/vcf_convert.html

Use this function to convert genotype matrices extracted from VCF files into the format required by pcadapt.

```R
vcf_convert(string_geno, output, allele_sep)
```

--------------------------------

### Perform PCA with LD Clumping

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/articles/pcadapt.html

Performs PCA after applying SNP thinning based on Linkage Disequilibrium (LD) clumping parameters (window size and r2 threshold). This helps to mitigate the effect of LD on PCA.

```R
res <- pcadapt(matrix, K = 20, LD.clumping = list(size = 200, thr = 0.1))
plot(res, option = "screeplot")
```

--------------------------------

### Retrieve population names function signature

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get.pop.names.html

Basic function signature for retrieving population names from a list of integers or strings.

```R
get.pop.names(pop)
```

--------------------------------

### get.score.color

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/get.score.color.html

Assigns colors to individuals based on their population for visualization purposes.

```APIDOC
## get.score.color

### Description
Allows the user to display individuals of the same predefined population with the same color when using the option "scores" in pcadapt.

### Parameters
#### Arguments
- **pop** (list of integers or strings) - Required - A list specifying which population the individuals belong to.

### Request Example
get.score.color(pop = c(1, 1, 2, 2, 3))
```

--------------------------------

### Generate Scree Plot for PCA Eigenvalues

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Visualize the proportion of variance explained by each principal component using a scree plot. You can specify the number of PCs to display.

```r
# Scree plot - proportion of variance explained by each PC
plot(x, option = "screeplot")
plot(x, option = "screeplot", K = 10)  # Show only first 10 PCs
```

--------------------------------

### vcf_convert Function

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/vcf_convert.html

Converts genotype matrices extracted from VCF files using vcfR to the format required by pcadapt.

```APIDOC
## vcf_convert

### Description
`vcf_convert` converts outputs of `extract.gt` to the format `pcadapt`.

### Method
R Function

### Endpoint
N/A (R function)

### Parameters
#### Arguments
- **string_geno** (matrix) - Required - a genotype matrix extracted from a VCF file with `vcfR`.
- **output** (character string) - Required - a character string indicating the name of the output file.
- **allele_sep** (vector of characters) - Required - a vector of characters indicating what delimiters are used to separate alleles.

### Request Example
```R
vcf_convert(string_geno = my_genotype_matrix, output = "output_file.geno", allele_sep = c("/"))
```

### Response
N/A (This function writes to a file and does not return a value in the typical API sense.)
```

--------------------------------

### vcf2pcadapt Function

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/vcf2pcadapt.html

The vcf2pcadapt function extracts genotype information from a VCF file and exports it in the format required by pcadapt. It utilizes the vcfR package for VCF file processing. Please note that this converter is not actively maintained, and the use of PLINK is recommended for new projects.

```APIDOC
## vcf2pcadapt

### Description
Converts a VCF file to the pcadapt format using the vcfR package.

### Method
Function call (R)

### Endpoint
N/A (R function)

### Parameters
#### Arguments
- **input** (character string) - Required - The name of the VCF file to be converted.
- **output** (character string) - Optional - The name of the output file. Defaults to "tmp.pcadapt".
- **allele.sep** (vector of characters) - Optional - Delimiters used to separate alleles. Defaults to c("\/", "|").

### Request Example
```R
vcf2pcadapt(input = "my_genotypes.vcf", output = "my_pcadapt_data.txt", allele.sep = c("/"))
```

### Response
#### Success Response
N/A (Function writes to a file)

#### Response Example
N/A
```

--------------------------------

### Display a Manhattan plot with manhattan_plot

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/manhattan_plot.html

Generates a Manhattan plot from a pcadapt object. Specify plt.pkg as 'plotly' for interactive plots or 'ggplot' for static ones.

```R
manhattan_plot(x, chr.info, snp.info, plt.pkg = "ggplot", K = 1)
```

--------------------------------

### Plot p-values Q-Q Plot

Source: https://github.com/bcm-uga/pcadapt/blob/master/docs/reference/qq_plot.html

Use `qq_plot` to visualize the distribution of computed p-values. Requires an output from the `outlier` function. The `K` argument specifies the principal component to display.

```R
qq_plot(x, K = 1)
```

--------------------------------

### LD Clumping Workflow in pcadapt

Source: https://context7.com/bcm-uga/pcadapt/llms.txt

Applies LD clumping to thin SNPs based on correlation thresholds before computing principal components. This helps mitigate spurious signals caused by linkage disequilibrium in dense SNP data.

```R
library(pcadapt)

# Load example data with LD structure
path_to_file <- system.file("extdata", "SSMPG2017.rds", package = "pcadapt")
genotypes <- readRDS(path_to_file)
matrix_input <- read.pcadapt(genotypes, type = "pcadapt")

# First, analyze without LD clumping to see the problem
res_no_ld <- pcadapt(matrix_input, K = 20)
plot(res_no_ld, option = "screeplot")  # May suggest incorrect K

# Check loadings for clustering (sign of LD artifacts)
par(mfrow = c(2, 2))
for (i in 1:4) {
  plot(res_no_ld$loadings[, i], pch = 19, cex = 0.3,
       ylab = paste0("Loadings PC", i))
}
# If loadings cluster in genomic regions, LD is affecting PCs

# Apply LD clumping: size = window radius, thr = r^2 threshold
res_ld <- pcadapt(matrix_input, K = 20,
                  LD.clumping = list(size = 200, thr = 0.1))
plot(res_ld, option = "screeplot")  # Now shows correct population structure

# Run final analysis with optimal K and LD clumping
res_final <- pcadapt(matrix_input, K = 2,
                     LD.clumping = list(size = 200, thr = 0.1))
```