CARNIVAL inputs
CARNIVAL requires at least two of input files which contains the information on:
- Prior Knowledge Network [PKN] i.e. "netFile"
- Measurement data i.e. "MeasFile"
Optionally, users can also provide additional information on:
- List for target(s) of perturbation i.e. "inputFile"
- (Pathway) Scoring data i.e. "weightFile".
Details information of each input files as provided as follows:
This file should contain 3 columns with a header line e.g. "source","Interaction","target". Each line below the header line refers to each interaction in the prior knowledge network. In the middle column, 1 refers to activatory interaction while -1 refers to inhibitory interactions. Note that all entries in each row are separated by tabs.
This file should contain 2 rows. The first row contains the name of all measured node(s) while the second row contains the value of the measurements. The measurement values can be either continuous or integer/discretised values, depends on how detailed measurement values should be taken into account in the CARNIVAL pipeline. The sign of the measurement value determines whether the respective measurements are up- or down-regulated. Note that all entries in each row are separated by tabs.
The measurement values as provided in the measFile will be applied as the multiplying factor of the mismatch penalty. For more information, please refer to the Supplementary Method section of the CARNIVAL publication.
This file should contain 2 rows. The first row contains the name of all input node(s) while the second row contains the state of the input node. Four input states are possible to be assigned i.e. [-1,0,1,NaN] for down-regulated, unperturbed (not recommended to be included), up-regulated, and unknown.
In case the state of an input node is set to NaN, the CARNIVAL pipeline will determine the most coherent state of the respective input node that matches the measurement data and it would be varied in each individual solution.
This file should contain 2 rows. The first row contains the names of all "pathways" for contextualising the network while the second row contains the "scores" of each "pathways". In the CARNIVAL pipeline, weightFile represents the signalling pathways scores from "PROGENy" which infers the activities of 14 pathways based on perturbation experiments.
The weight/score values as provided in the weightFile will be applied as the soft penalising factor of the size penalty (depended on the number of nodes). The implication of the penalty will be applied on the representative node(s) of each pathway (stored in "progenyMembers.RData" in the CARNIVAL package). For more information, please refer to the Supplementary Method section of the CARNIVAL publication.