Features of lotus 12311/6/2022 "Kernel_similarity.RData" is only necessary for LOTUS2. Task <- 1 driver_type <- 'tsg ' # 'tsg' or 'og' version <- 'lotus ' # 'lotus' for LOTUS, 'lotus2' for LOTUS2, 'aggregation' for aggregation LOTUS and 'onetask' for onetask LOTUSįor example, when task is 1, driver_type is "tsg" and version is "lotus", run_lotus_multitask.R computes a ranking of all genes listed in features_2020.txt (which is the default dataset), except for the known type-specific tumor suppressors in tsg_per_Diseases.RData, according to their potential as tumor suppressors, using LOTUS with a training set of known tumor suppressors from tsg_per_Diseases.RData. Then, choose the arguments in the script: #Features of lotus 123 driver#To run the script "run_lotus_multitask.R", you need at a minimum the source file "lotus.R", a feature matrix (for example "features_2020.txt"), a corresponding PPI kernel (for example "PPIKernel_2020.txt"), a similarity matrix between diseases (for example "Kernel_similarity.RData") and lists of known cancer driver genes for every considered disease (for example "tsg_per_Diseases.RData"). This script returns the consistency error of LOTUS trained with the chosen data. Then, choose the arguments in the script as for "run_lotus.R". To run the script "evaluate_lotus.R", you need at a minimum the source files "lotus.R" and "evaluation.R", a feature matrix (for example "features_2020.txt") and a corresponding PPI kernel (for example "PPIKernel_2020.txt"). Alternatively, for example if you wish to use other features, you will have to change the corresponding lines in the script. Pay attention to the fact that the feature matrix should contain columns named "Frameshift", "LOF", "Splice" for tumor suppressor prediction, and "Entropy.Score", "Missense.Damaging", "Missense.total" for oncogene prediction. #Features of lotus 123 code#To use LOTUS with different datasets, you only have to change the lines where the features, the set of known driver genes and possibly the PPI kernel so that the code fits the name of your datasets. Driver_type <- 'tsg ' # 'tsg' or 'og' dataset <- 'cosmicv86 ' # '2020', 'cosmicv86', 'mutsig' or 'tuson' ppi <- 'yes ' # 'yes', 'no' or 'only'įor example, when driver_type is "tsg", dataset is "tuson" and ppi is "yes", run_lotus.R computes a ranking of all genes listed in features_tuson.txt, except for the known tumor suppressors in tsg_tuson.txt, according to their potential as tumor suppressors, using a training set of known tumor suppressors from tsg_tuson.txt and information from the PPI network.
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |