MDT-class: Mutation Data Table
In olivmrtn/MachineLearningGWAS: Machine Learning for Genome Wide Association Studies

Description Arguments Slots Constructors Accessors Subset Plots Add Annotations Train Classifier See Also

The MDT class can contain information about sequence variation, annotations and phenotype for a collection of samples. It is meant to simplify machine learning in genome-wide association studies.

`x`	`MDT` object.
`value`	`character` vector.

mtable: Variation matrix. Must contain the following columns: FeatureID, SampleID, VALUE.
annotations: Feature annotations. Must contain the following columns: FeatureID.
phenotype: Sample annotations. Must contain the following columns: SampleID, RESPONSE (variable to be predicted), SEX, AGE.
info: Feature and Sample annotations. Annotations that are dependant on both sample and feature such as quality measures or observed alleles. Must contain the following columns: FeatureID, SampleID.
features: character vector of unique FeatureIDs. Each combination of chromosome and position makes one unique FeatureID.
samples: character vector of unique SampleIDs.

MDT contructs a MDT object. Only the mtable argument is required.

vcfsToMDT converts a list of VCF objects to MDT without any phenotypic data, that is, with an empty phenotype slot.

importPhenotype adds phenotypic data to MDT object.

aggregateMDT aggregate FeatureIDs Into Higher Level IDs.

mtable(x), mtable(x) <- value gets or sets mtable.

annotations(x), annotations(x) <- value gets or sets annotations.

phenotype(x), phenotype(x) <- value gets or sets phenotype.

info(x), info(x) <- value gets or sets info.

response(x), response(x) <- value gets or sets Response column in phenotype slot.

features(x), features(x) <- value gets or sets FeatureIDs.

samples(x), samples(x) <- value gets or sets SampleIDs.

asMatrixMDT(x, miss) converts mtable tp a matrix with SampleID as rows and FeatureIDs as columns.

x[SampleIDs, FeatureIDs] subsets MDT objects using SampleIDs and FeatureIDs. Must be character.

filterMDT filter samples and features in MDT objects using annotations in annotations, phenotype and info.

plotMDT allows to generate various plots such as PCA, heatmaps, histograms.

manhattanPlot creates a Manhattan plot.

See annotateMDT to add feature annotations to annotations(x).

See importPhenotype to add sample annotations to phenotype(x).

Use selectMDT to retrieve annotations from annotations(x), phenotyope(x) and info(x).

trainClassifier trains a classifier, using an MDT object, that that predicts phenotypic response(x) using mtable(x) matrix. Creates a MLGWAS object.

MLGWAS data.table

olivmrtn/MachineLearningGWAS documentation built on May 24, 2019, 12:52 p.m.