The MIPPIE rescoring tool (rescore_mippie) takes as an input a flat file version of the MIPPIE database and a list of scores assigned to experimental techniques. rescore_mippie is available at http://cbdm-01.zdv.uni-mainz.de/~galanisl/mippie/RS/rescore_mippie.R Requirements and installation ============================= The script is written in R and requires: - R >= v3.5.1 - readr >= 1.1.1 - dplyr >= 0.7.6 No further installation is necessary, the tool can be run right away from the R console. It reads the MIPPIE interaction network from a flat file and the scores assigned to the experimental techniques used to measure these interactions. The up-to-date versions of these files can be found at http://cbdm-01.zdv.uni-mainz.de/~galanisl/mippie/downloads/mippie_ppi_v1_0.tsv http://cbdm-01.zdv.uni-mainz.de/~galanisl/mippie/RS/experimental_scores.tsv By default rescore_mippie uses a flat file called mippie_ppi_v1_0.tsv that is expected to reside in the same directory as the script itself. Quick Start =========== In the following, the dollar sign ($) corresponds to the terminal's prompt and the greater than symbol (>) corresponds to R's prompt. From the terminal, change to the directory containing rescore_mippie.R. For example, if it is in folder Downloads do: $ cd /home/user_name/Downloads Start R: $ R Once inside an R session, source rescore_mippie.R to make its main function available: > source("rescore_mippie.R") When MIPPIE and the list of experimental scores are in the working directory, the tool can be run without passing any further parameters: > mippie_rs <- rescore_mippie() The above both generates a file called mippie_ppi_v1_0_rescored.tsv in the working directory and returns a data frame for further post-processing in R. The name of the output file can be changed as follows: > mippie_rs <- rescore_mippie(output.file = "mippie_rescored.tsv") If MIPPIE or the list of experimental scores are in a different directory, their path must be specified: > mippie_rs <- rescore_mippie(path.to.mippie = "/home/user_name/Docs/mippie_ppi_v1_0.tsv", path.to.exp.scores = "/home/user_name/Docs/experimental_scores.tsv") A rescoring run putting a higher emphasis on the orthology subscore but leaving the saturation parameters as they are (for details on the scoring formula please see the paper describing MIPPIE) looks like: > mippie_rs <- rescore_mippie(w_s = 0.5, w_o = 0.3, w_t = 0.2) Parameters for the rescore_mippie function: ======= path.to.mippie character; The path to the file containing the MIPPIE path.to.exp.scores character; The path to the file reporting the scores assigned to the experimental techniques used to measure MIPPIE's interactions. a_s numeric; Controls how fast the study subscore saturates (between 0 and Inf), default value (as used in the paper) is 1. a_o numeric; Controls how fast the orthology subscore saturates (between 0 and Inf), default value (as used in the paper) is 1.5. a_t numeric; Controls how fast the technique subscore saturates (between 0 and Inf), default value (as used in the paper) is 0.5. w_s numeric; Contribution of the study subscore to the overall score (between 0 and 1), default value (as used in the paper) is 0.6. w_o numeric; Contribution of the orthology subscore to the overall score (between 0 and 1), default value (as used in the paper) is 0.1. w_t numeric; Contribution of the technique subscore to the overall score (between 0 and 1), default value (as used in the paper) is 0.3. output.file character; Path and name of the output files, which will contain the re-scored version of MIPPIE. For any questions, comments or suggestions, please send an email to galanisl@uni-mainz.de