featureMatrix: Form a feature matrix using lists of positive and negative...
In ShaoxunLiu/proteinStructureBoost: Generate Boosting Model for Structured Protein

This function takes a vector of positively labeled pdb file names and a negatively labeled pdb file names to generate a feature matrix for machine learning model formation. pbd files can be randomly rotated to increase sample size.

featureMatrix(
  positive,
  negative = NA,
  projectFile = "~/",
  as.file = TRUE,
  balance = FALSE,
  positive.label = 1,
  negative.label = 0,
  rotations = 0,
  size = 32,
  units = 16,
  combine.chains = TRUE,
  show.grid = FALSE,
  n.feature = 68
)

`positive`	Character vector of positively labeled pdb files.
`negative`	Character vector of negatively labeled pdb files.
`projectFile`	The directory in which the PDB files are downloaded.
`as.file`	If true, creates a /PDBs folder to store downloaded files.
`balance`	If true balances the number of positive and negative samples.
`positive.label`	Label assigned for positive samples.
`negative.label`	Label assigned for negative samples.
`rotations`	Number of rotational augmentation samples added to the matrix.
`size`	The length of the recognition cube in Angstroms
`units`	The number of blocks in each axis.
`combine.chains`	If true, assign all chains in the pdb file in one grid. If false, generates a separate grid for each chain.
`show.grid`	If true, outputs an .html file that contains the 3D representation of the grid. Directory to output 3D representation .html file is the same as the .pdb file.
`n.feature`	Suggested number of features in the feature matrix. Will be fixed to multiples of 68 (a result of lbp3D algorithm).
`center.median`	If true, centers each chain (or all chains depending on combine.chains) to the median of their xyz.