In Science this week, researchers from the California Institute of Technology report data showing that variability in gene expression between genetically identical cells arises in predictable ways that are based on promoter architecture. The team created a set of promoters in Escherichia coli in which promoter strength, transcription factor binding strength, and transcription factor copy numbers were systematically varied, and used mRNA fluorescence in situ hybridization to watch how these differences affected gene expression. A comparison of the results against model predictions without parameters revealed that the "molecular details of transcription dictate variability in mRNA expression," and that transcriptional noise is tunable.
Also in Science, researchers from the University of Toronto detail a new computational model that can predict how strongly difference genetic variants affect RNA splicing, which is mutated in a variety of diseases. They analyzed more than 650,000 genetic variants in both coding and non-coding regions of the human genome and found widespread patterns of how these mutations can cause splicing errors. The model was able to identify many known splicing mutations in a variety of diseases, as well as several new genes potentially related to autism.