This analysis comprehensively compares methods for gene regulatory network inference submitted through the DREAM5 challenge. It demonstrates that integration of predictions from multiple methods shows the most robust performance across data sets. Reconstructing gene regulatory networks from high-throughput data is a long-standing challenge. Through the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we performed a comprehensive blind assessment of over 30 network inference methods on Escherichia coli, Staphylococcus aureus, Saccharomyces cerevisiae and in silico microarray data. We characterize the performance, data requirements and inherent biases of different inference approaches, and we provide guidelines for algorithm application and development. We observed that no single inference method performs optimally across all data sets. In contrast, integration of predictions from multiple inference methods shows robust and high performance across diverse data sets. We thereby constructed high-confidence networks for E. coli and S. aureus, each comprising ∼1,700 transcriptional interactions at a precision of ∼50%. We experimentally tested 53 previously unobserved regulatory interactions in E. coli, of which 23 (43%) were supported. Our results establish community-based methods as a powerful and robust tool for the inference of transcriptional gene regulatory networks.

A community of researchers report the lessons learned from applying 44 algorithms to predict drug sensitivity in cancer cell lines using genomic, epigenetic and proteomic datasets Predicting the best treatment strategy from genomic information is a core goal of precision medicine. Here we focus on predicting drug response based on a cohort of genomic, epigenomic and proteomic profiling data sets measured in human breast cancer cell lines. Through a collaborative effort between the National Cancer Institute (NCI) and the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we analyzed a total of 44 drug sensitivity prediction algorithms. The top-performing approaches modeled nonlinear relationships and incorporated biological pathway information. We found that gene expression microarrays consistently provided the best predictive power of the individual profiling data sets; however, performance was increased by including multiple, independent data sets. We discuss the innovations underlying the top-performing methodology, Bayesian multitask MKL, and we provide detailed descriptions of all methods. This study establishes benchmarks for drug sensitivity prediction and identifies approaches that can be leveraged for the development of new methods.

Venetoclax-based therapy can induce responses in approximately 70% of older previously untreated patients with acute myeloid leukemia (AML). However, up-front resistance as well as relapse following initial response demonstrates the need for a deeper understanding of resistance mechanisms. In the present study, we report that responses to venetoclax +azacitidine in patients with AML correlate closely with developmental stage, where phenotypically primitive AML is sensitive, but monocytic AML is more resistant. Mechanistically, resistant monocytic AML has a distinct transcriptomic profile, loses expression of venetoclax target BCL2, and relies on MCL1 to mediate oxidative phosphorylation and survival. This differential sensitivity drives a selective process in patients which favors the outgrowth of monocytic subpopulations at relapse. Based on these findings, we conclude that resistance to venetoclax + azacitidine can arise due to biological properties intrinsic to monocytic differentiation. We propose that optimal AML therapies should be designed so as to independently target AML subclones that may arise at differing stages of pathogenesis. Significance: Identifying characteristics of patients who respond poorly to venetoclax-based therapy and devising alternative therapeutic strategies for such patients are important topics in AML. We show that venetoclax resistance can arise due to intrinsic molecular/metabolic properties of monocytic AML cells and that such properties can potentially be targeted with alternative strategies.

Prolonged antibiotic treatment can lead to detrimental side effects in patients, including ototoxicity, nephrotoxicity, and tendinopathy, yet the mechanisms underlying the effects of antibiotics in mammalian systems remain unclear. It has been suggested that bactericidal antibiotics induce the formation of toxic reactive oxygen species (ROS) in bacteria. We show that clinically relevant doses of bactericidal antibiotics—quinolones, aminoglycosides, and β-lactams—cause mitochondrial dysfunction and ROS overproduction in mammalian cells. We demonstrate that these bactericidal antibiotic–induced effects lead to oxidative damage to DNA, proteins, and membrane lipids. Mice treated with bactericidal antibiotics exhibited elevated oxidative stress markers in the blood, oxidative tissue damage, and up-regulated expression of key genes involved in antioxidant defense mechanisms, which points to the potential physiological relevance of these antibiotic effects. The deleterious effects of bactericidal antibiotics were alleviated in cell culture and in mice by the administration of the antioxidant N-acetyl-l-cysteine or prevented by preferential use of bacteriostatic antibiotics. This work highlights the role of antibiotics in the production of oxidative tissue damage in mammalian cells and presents strategies to mitigate or prevent the resulting damage, with the goal of improving the safety of antibiotic treatment in people.