Research models of neurodevelopmental disorders: The right model in the right place.

Neurodevelopmental disorders (NDDs) are a heterogeneous group of impairments that affect the development of the central nervous system leading to abnormal brain function. NDDs affect a great percentage of the population worldwide, imposing a high societal and economic burden and thus, interest in this field has widely grown in recent years. Nevertheless, the complexity of human brain development and function as well as the limitations regarding human tissue usage make their modeling challenging. Animal models play a central role in the investigation of the implicated molecular and cellular mechanisms, however many of them display key differences regarding human phenotype and in many cases, they partially or completely fail to recapitulate them. Although in vitro two-dimensional (2D) human-specific models have been highly used to address some of these limitations, they lack crucial features such as complexity and heterogeneity. In this review, we will discuss the advantages, limitations and future applications of in vivo and in vitro models that are used today to model NDDs. Additionally, we will describe the recent development of 3-dimensional brain (3D) organoids which offer a promising approach as human-specific in vitro models to decipher these complex disorders.

A Deep Learning Framework for Predicting Response to Therapy in Cancer.

A major challenge in cancer treatment is predicting clinical response to anti-cancer drugs on a personalized basis. Using a pharmacogenomics database of 1,001 cancer cell lines, we trained deep neural networks for prediction of drug response and assessed their performance on multiple clinical cohorts. We demonstrate that deep neural networks outperform the current state in machine learning frameworks. We provide a proof of concept for the use of deep neural network-based frameworks to aid precision oncology strategies.

GemC1 is a critical switch for neural stem cell generation in the postnatal brain.

The subventricular zone (SVZ) is one of two main niches where neurogenesis persists during adulthood, as it retains neural stem cells (NSCs) with self-renewal capacity and multi-lineage potency. Another critical cellular component of the niche is the population of postmitotic multiciliated ependymal cells. Both cell types are derived from radial glial cells that become specified to each lineage during embryogenesis. We show here that GemC1, encoding Geminin coiled-coil domain-containing protein 1, is associated with congenital hydrocephalus in humans and mice. Our results show that GemC1 deficiency drives cells toward a NSC phenotype, at the expense of multiciliated ependymal cell generation. The increased number of NSCs is accompanied by increased levels of proliferation and neurogenesis in the postnatal SVZ. Finally, GemC1-knockout cells display altered chromatin organization at multiple loci, further supporting a NSC identity. Together, these findings suggest that GemC1 regulates the balance between NSC generation and ependymal cell differentiation, with implications for the pathogenesis of human congenital hydrocephalus.