Extended Data Fig. 4: Replication of a TCGA DLVPM model in CPTAC data.
From: Integrating multimodal cancer data using deep latent variable path modelling
a: For each data type, these plots show the mean Pearson’s correlation of each DLV, with DLVs from data types connected by the path model, in the CPTAC dataset (n = 105). The error bars on the plot denote mean-centred 95% bootstrapped confidence intervals. b: Association matrices for all five DLVs. Entries in the upper triangular part of the matrix indicate Pearson’s correlation values between different data-types. Entries in the lower part of the matrix are significance values for these correlations, obtained using permutation testing.
