Skip to contents

Rank drugs by how strongly they reverse the meta-signature

Source: R/09_downstream.R
mx_drug_connectivity.Rd

Scores each drug perturbation signature (sets of genes the drug pushes up and down) against the meta-analysis ranking. The reversal score is a standardised difference of the mean signed meta z-score of the drug's up and down genes; strongly negative values mean the drug moves genes in the opposite direction to the disease and are candidate therapeutics. The Connectivity Map Kolmogorov-Smirnov score (Lamb et al. 2006) is reported alongside for comparison.

Usage

mx_drug_connectivity(
  meta_result,
  signatures,
  id_col = "gene_id",
  min_size = 10
)

Arguments

meta_result

A metaXpressResult object.

signatures

A named list of drug signatures, each a list with character vectors up and down, as returned by mx_read_signatures.

id_col

Column of meta_table whose values match the signature genes. Default "gene_id".

min_size

Minimum number of measured genes in each of the up and down sets. Default 10.

Value

A data.frame sorted by reversal score with columns drug, reversal_score (negative = reverses the disease signature), pvalue, padj, cmap_score, n_up, n_down. P-values assume independent genes and are best used for ranking rather than as calibrated error rates.

References

Lamb, J. et al. (2006) The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease. Science, 313(5795), 1929–1935. doi:10.1126/science.1132939

Subramanian, A. et al. (2017) A next generation Connectivity Map: L1000 platform and the first 1,000,000 profiles. Cell, 171(6), 1437–1452. doi:10.1016/j.cell.2017.10.049

See also

mx_read_signatures, mx_downstream_plot

Examples

mt <- data.frame(gene_id = paste0("G", 1:300),
                 meta_log2FC = c(rep(2, 30), rep(-2, 30), rnorm(240, sd = 0.2)),
                 meta_pvalue = c(rep(1e-8, 60), runif(240)),
                 meta_padj = 1, i_squared = 0, n_studies = 3L)
res <- new("metaXpressResult", meta_table = mt, method = "fisher",
           n_studies = 3L)
sigs <- list(reverser = list(up = paste0("G", 31:60), down = paste0("G", 1:30)),
             mimic    = list(up = paste0("G", 1:30), down = paste0("G", 31:60)))
mx_drug_connectivity(res, sigs)
#>       drug reversal_score       pvalue         padj cmap_score n_up n_down
#> 1 reverser      -16.29959 9.935399e-60 9.935399e-60  -1.803333   30     30
#> 2    mimic       16.29959 9.935399e-60 9.935399e-60   1.803333   30     30