clustMetrIterTiMEx: Assesses the performance of the inferred groups after running...

In cbg-ethz/pathTiMEx: Joint Inference of Mutually Exclusive Driver Pathways and their Progression Dynamics

Description

clustMetrIterTiMEx computes per-edge and per-pathway metrics evaluating the performance of the pathway assignment inferred after running TiMEx iteratively, based on data simulated from PathTiMEx

Usage

clustMetrIterTiMEx(resIterTiMEx)

Arguments

resIterTiMEx

structure containing the results from iteratively running TiMEx on a binary dataset, as returned by the function iterTiMExSims.

Details

This function assesses how well the real groups were recovered after performing iterative inference with TiMEx.

Value

list consisting of

• reconstruction structure with metrics of the reconstuction of the initial groups, as follows

  • groupsSpreadlist containing as many elements as real pathways. each element is a list with at most as many elements as the number of inferred pathways (including potential NULL elements). each element of this second list is a vector of genes part of the current real pathway and also part of the inferred pathway. In othder words, the first index represents the real group and the second index represents the inferred group.

  • groupsSpreadPercent list containing as many elements as real pathways. each element is a list with at most as many elements as the number of inferred pathways (including potential NULL elements). each element of this second list is the percentage of reconstruction of the respective first pathway by the genes identified as part of the second pathway. In othder words, the first index represents the real group and the second index represents the inferred group.

  • groupsSpreadOrd list containing as many elements as real pathways, ordered by the real mu values of the pathways. each element is a list with at most as many elements as the number of inferred pathways (including potential NULL elements). each element of this second list is a vector of genes part of that particular real pathway and also inferred as part of the current pathway. In othder words, the first index represents the real group and the second index represents the inferred group.

  • groupsSpreadPercentOrd list containing as many elements as real pathways, orderde by the real mu values of the pathways. each element is a list with at most as many elements as the number of inferred pathways (including potential NULL elements). each element of this second list is the percentage of reconstruction of the respective first pathway by the genes identified as part of the second pathway. In othder words, the first index represents the real group and the second index represents the inferred group.

  • groupsCompact list containing as many elements as real pathways. each element is a vector consisting of the genes members of that particular pathway which were identified as part of a mutually exclusive group, regardless of the group in which they were identified.

  • groupsCompactPercent list containing as many elements as real pathways. each element is the percentage of reconstruction of that particular pathway, i.e. the percentage of genes which were identified as mutually exclusive, regardless of the group in which they were identified.

  • groupsCompactOrd list containing as many elements as real pathways, ordered by the real mu values of the pathways. each element is a vector consisting of the genes members of that particular pathway which were identified as part of a mutually exclusive group, regardless of the group in which they were identified.

  • groupsCompactPercentOrd list containing as many elements as real pathways, ordered by the real mu values of the pathwayss. each element is the percentage of reconstruction of that particular pathway, i.e. the percentage of genes which were identified as mutually exclusive, regardless of the group in which they were identified.

  • groupsEmpty list containing as many elements as real pathways. each element contains the genes from that particular pathway which were not identified as part of mutually exclusive groups.

  • groupsEmptyPercent list containing as many elements as real pathways. each element is the percentage of the respective pathway which was not identified as part of mutually exclusive groups.

  • groupsEmptyOrd list containing as many elements as real pathways, ordered by the real mu values of the pathways. each element contains the genes from that particular pathway which were not identified as mutually exclusive.

  • groupsEmptyPercentOrd list containing as many elements as real pathways, ordered by the real mu values of the pathways. each element is the percentage of the respective pathway which was not identified as part of mutually exclusive groups.

  • booleanOthers list containing as many elements as real pathways. each element is a list with at most as many elements as the number of inferred pathways (including potential NULL elements). each element of this second list is either 0, if the inferred group contained elements from more than one real pathway, or 1, if the inferred group only contained elements from one real pathway. In othder words, the first index represents the real group and the second index represents the inferred group.

  • booleanOthersOrdlist containing as many elements as real pathways, orderde by the real mu values of the pathways. each element is a list with at most as many elements as the number of inferred pathways (including potential NULL elements). each element of this second list is either 0, if the inferred group contained elements from more than one real pathway, or 1, if the inferred group only contained elements from one real pathway. In othder words, the first index represents the real group and the second index represents the inferred group.

• characteristics structure with metrics per pathway characterizing the identified groups, as follows

  • freqGenes list containing as many elements as real pathways. each element is a vector with frequencies of the genes inside that pathway.

  • freqGenesAvg list containing as many elements as real pathways. each element is the average frequency of all genes inside that pathway (averaged over all the genes in each pathway).

  • freqGenesOrd list containing as many elements as real pathways. each element is a vector with the frequencies of the genes inside that pathway, ordered decreasingly by the intensities of mutual exclusivity of each pathway.

  • realMus vector containing the intensities of mutual exclusivity of the real pathways

  • realMusOrd vector containing the intensities of mutual exclusivity of the real pathways, ordered decreasingly

• peredge structure with metrics per edge characterizing the reconstruction of groups, as follows

  • tp true positive rate: the number of true mutual exclusivity pairwise connections which were also identified, divided by the total number of true mutual exclusivity pairwise connections.

  • fp false positive rate: the number of false mutual exclusivity edges identified, divided by the total number of possible pairwise connections.

• resIterTiMEx input structure containing the identified mutually exclusive groups after running TiMEx itartively, together with the simulated data on the basis of which the inference was done.

Author(s)

Simona Constantinescu, [e-mail]

cbg-ethz/pathTiMEx documentation built on May 13, 2019, 2:03 p.m.