A single-cell transcriptomic atlas of complete insect nervous systems across multiple life stages.

Molecular profiles of neurons influence neural development and function but bridging the gap between genes, circuits, and behavior has been very difficult. Here we used single cell RNAseq to generate a complete gene expression atlas of the Drosophila larval central nervous system composed of 131,077 single cells across three developmental stages (1 h, 24 h and 48 h after hatching). We identify 67 distinct cell clusters based on the patterns of gene expression. These include 31 functional mature larval neuron clusters, 1 ring gland cluster, 8 glial clusters, 6 neural precursor clusters, and 13 developing immature adult neuron clusters. Some clusters are present across all stages of larval development, while others are stage specific (such as developing adult neurons). We identify genes that are differentially expressed in each cluster, as well as genes that are differentially expressed at distinct stages of larval life. These differentially expressed genes provide promising candidates for regulating the function of specific neuronal and glial types in the larval nervous system, or the specification and differentiation of adult neurons. The cell transcriptome Atlas of the Drosophila larval nervous system is a valuable resource for developmental biology and systems neuroscience and provides a basis for elucidating how genes regulate neural development and function.

Clustering of Drosophila housekeeping promoters facilitates their expression.

Housekeeping genes of animal genomes cluster in the same chromosomal regions. It has long been suggested that this organization contributes to their steady expression across all the tissues of the organism. Here, we show that the activity of Drosophila housekeeping gene promoters depends on the expression of their neighbors. By measuring the expression of ∼85,000 reporters integrated in Kc167 cells, we identified the best predictors of expression as chromosomal contacts with the promoters and terminators of active genes. Surprisingly, the chromatin composition at the insertion site and the contacts with enhancers were less informative. These results are substantiated by the existence of genomic "paradoxical" domains, rich in euchromatic features and enhancers, but where the reporters are expressed at low level, concomitant with a deficit of interactions with promoters and terminators. This indicates that the proper function of housekeeping genes relies not on contacts with long distance enhancers but on spatial clustering. Overall, our results suggest that spatial proximity between genes increases their expression and that the linear architecture of the Drosophila genome contributes to this effect.

Machine Learning: How Much Does It Tell about Protein Folding Rates?

The prediction of protein folding rates is a necessary step towards understanding the principles of protein folding. Due to the increasing amount of experimental data, numerous protein folding models and predictors of protein folding rates have been developed in the last decade. The problem has also attracted the attention of scientists from computational fields, which led to the publication of several machine learning-based models to predict the rate of protein folding. Some of them claim to predict the logarithm of protein folding rate with an accuracy greater than 90%. However, there are reasons to believe that such claims are exaggerated due to large fluctuations and overfitting of the estimates. When we confronted three selected published models with new data, we found a much lower predictive power than reported in the original publications. Overly optimistic predictive powers appear from violations of the basic principles of machine-learning. We highlight common misconceptions in the studies claiming excessive predictive power and propose to use learning curves as a safeguard against those mistakes. As an example, we show that the current amount of experimental data is insufficient to build a linear predictor of logarithms of folding rates based on protein amino acid composition.