Genes are ranked by decreasing correlation with the phenotype.
Walking down the ranked list, a running sum increases by
\(|r_j|^p / N_R\) (with \(N_R = \sum_{g_j \in S} |r_j|^p\))
when gene j belongs to the set S and decreases by
\(1/(N - N_H)\) otherwise. The enrichment score ES is the
maximum deviation from zero of the running sum, a weighted
Kolmogorov-Smirnov-like statistic; p = 0 reduces to the standard
Kolmogorov-Smirnov statistic and p = 1 is the paper default. With
nperm > 0 a nominal p-value is estimated by permuting the
gene labels of the set (pre-ranked variant), assessing positive
and negative scores separately against the same-sign side of the
permutation null as prescribed in the paper.
Value
List with es, arg_es (0-based rank of the
extreme deviation), running, n, n_hits,
nperm, pvalue (when nperm > 0),
method.
References
Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette, M. A., Paulovich, A., Pomeroy, S. L., Golub, T. R., Lander, E. S. and Mesirov, J. P. (2005), Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles, PNAS 102(43), 15545-15550. Appendix, Enrichment Score ES(S), p. 15550; sign-separated significance, p. 15546.
Examples
set.seed(1)
Gnsetenr(correlations = rnorm(20), in_set = c(rep(TRUE, 5), rep(FALSE, 15)))
#> $es
#> [1] 0.446793
#>
#> $arg_es
#> [1] 0
#>
#> $running
#> [1] 4.467930e-01 3.801263e-01 3.134597e-01 2.467930e-01 1.801263e-01
#> [6] 1.134597e-01 4.679300e-02 -1.987367e-02 -8.654033e-02 -1.532070e-01
#> [11] -6.092120e-02 -9.487903e-03 -7.615457e-02 -1.428212e-01 -2.094879e-01
#> [16] -2.761546e-01 -1.007026e-01 -1.673693e-01 6.666667e-02 6.938894e-17
#>
#> $n
#> [1] 20
#>
#> $n_hits
#> [1] 5
#>
#> $nperm
#> [1] 0
#>
#> $method
#> [1] "GSEA enrichment score (Subramanian et al. 2005)"
#>