Description Usage Arguments Details Value Examples

Integrate two omics data type (and a quantitative phenotype), and calculate the absolute canonical correlation weights for the omics features using SmCCA SsCCA, or SCCA. SmCCA and SsCCA take into account a phenotype/trait. SmCCA maximizes the total (weighted or unweighted) pairwise canonical correlation weights between two omics data types and the trait. It requires the trait to be quantitative. SsCCA prioritizes omics features based on the trait, and assigns non-zero canonical weights to features that are more correlated to the trait. SCCA does not use any trait information for computing the canonical correlation weights. All of these three methods are included in this function, along with an omics feature subsampling scheme.

1 2 3 |

`X1` |
An |

`X2` |
An |

`Trait` |
An |

`Lambda1` |
LASSO penalty parameter for |

`Lambda2` |
LASSO penalty parameter for |

`s1` |
Proportion of mRNA features to be included, default at |

`s2` |
Proportion of miRNA features to be included, default at |

`NoTrait` |
Logical, default is |

`FilterByTrait` |
Logical, default is |

`SubsamplingNum` |
Number of feature subsamples. Default is 1000. Larger number leads to more accurate results, but at a higher cost. |

`CCcoef` |
Optional coefficients for the SmCCA pairwise canonical
correlations. If |

`trace` |
Logical. Whether to display the CCA algorithm trace. |

To choose SmCCA, set `NoTrait = FALSE, FilterByTrait = FALSE`

.
To choose SsCCA, set `NoTrait = FALSE, FilterByTrait = TRUE`

.
To choose SCCA, set `Trait = NULL, NoTrait = TRUE`

.

A canonical correlation weight matrix with *p_1+p_2* rows. Each
column is the canonical correlation weights based on subsampled `X1`

and `X2`

features. The number of columns is `SubsamplingNum`

.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | ```
## For illustration, we only subsample 5 times.
set.seed(123)
# Unweighted SmCCA
W1 <- getRobustPseudoWeights(geneExpr, mirnaExpr, Trait = pheno, Lambda1 = 0.05,
Lambda2 = 0.05, s1 = 0.7, s2 = 0.9, NoTrait = FALSE, FilterByTrait = FALSE,
SubsamplingNum = 5, CCcoef = NULL, trace = FALSE)
# Weighted SmCCA
W2 <- getRobustPseudoWeights(geneExpr, mirnaExpr, Trait = pheno, Lambda1 = 0.05,
Lambda2 = 0.05, s1 = 0.7, s2 = 0.9, NoTrait = FALSE, FilterByTrait = FALSE,
SubsamplingNum = 5, CCcoef = c(1, 5, 5), trace = FALSE)
# SsCCA
W3 <- getRobustPseudoWeights(geneExpr, mirnaExpr, Trait = pheno, Lambda1 = .05, Lambda2 = 0.5,
s1 = 0.7, s2 = 0.9, NoTrait = FALSE, FilterByTrait = TRUE,
SubsamplingNum = 5, CCcoef = NULL, trace = FALSE)
# SCCA
W4 <- getRobustPseudoWeights(geneExpr, mirnaExpr, Trait = NULL, Lambda1 = 0.05,
Lambda2 = 0.05, s1 = 0.7, s2 = 0.9, NoTrait = TRUE,
SubsamplingNum = 5, CCcoef = NULL, trace = FALSE)
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.