Skip to content

4. Screenshots

Giulio Caravagna edited this page Jul 2, 2018 · 13 revisions

We used the pio package (Pretty I/O) to print coloured console output. pio, in turns, exploits the crayon package; coloured output helps a lot spotting immediate relevant information from REVOLVER's output.

  • to enable crayon in RStudio you want to see this.
  • if your terminal does not support UNICODE, you can disable crayon via: options(crayon.enabled = FALSE). This will not affect pio's functions, and the output will not be coloured.

Here are some example outputs: click to view full size images.


Screen_Shot_2018_06_30_at_11_04_24

An input tree (phylogenetic or mutation tree): an S3 object of class rev_phylo. It describes a model for one of our patients, its structure, its score and the information transfer for the annotated drivers.


Screen_Shot_2018_06_30_at_11_17_50

A run of REVOLVER fit: its EM and the expansion for a simple cohort. For each patient the number of solutions (trees) is reported, as well as the different combinations of information transfer. An initial condition (a tree per patient) is selected, and the EM is executed (numbers are the rank of the selected solution for each patient (1 would be the solution selected without REVOLVER). In this case, in one step the EM converges because there the trees are so simple that there is not much information to correlated them. In a more complex scenario, red numbers refer to cases in which the MLE estimate in the M-step of the EM differs from the MLE; the algorithm converges when all estimates are locally optimal (all green numbers). At the end, computation of intra-group trajectories is carried out by transferring orderings across patients.


Screen_Shot_2018_06_30_at_11_28_40

A cohort after fitting is an object of class rev_cohort_fit. Once you print it, basic descriptions and stats are reported.

Clone this wiki locally