Gene network modeling via TopNet reveals functional dependencies between diverse tumor-critical mediator genes

Abstract

Highlights d A network of non-mutated genes is critical to the malignant state d TopNet can accurately model cellular responses to genetic perturbations d TopNet is capable of pinpointing key architectural features of cancer cells

Protocol to use TopNet for gene regulatory network modeling using gene expression data from perturbation experiments

“…These parameters should be chosen such that either the network reaches a score of zero (indicating a fit that perfectly explains the observed data) or until additional cycles do not produce a reduction in the network score. The code used to produce these network fits is shown below:

library("ternarynet") data("crgnet_scores") results <- parallelFit(experiment_set=crgnet_scores, max_parents=4, n_cycles=1e9, n_write=10, T_lo=0.001, T_hi=1, target_score=0, n_proc=20, logfile="tnet-fit.log"), seed=as.integer(112358) )

Note: The computational time required to generate these network models is substantial: each independent network fit presented in McMurray et al. (2021) took approximately 12 h to run in parallel on 20 compute nodes.…”

“…Note: The computational time required to generate these network models is substantial: each independent network fit presented in McMurray et al. (2021) took approximately 12 h to run in parallel on 20 compute nodes.…”

“… Animals should be implanted with only one type of cell population – vector control or a single gene perturbation per animal. For experiments in McMurray et al. (2021) , six implantations were done using cells from each perturbed cell population.…”

“…While developed for gene perturbation experiments, TopNet models data in which nodes are both perturbed and measured. For complete details on the use and execution of this protocol, please refer to McMurray et al. (2021) .…”

“…For complete details on the use and execution of this protocol, please refer to McMurray et al. (2021) .…”