Search

Self-driving labs (SDLs) combine fully automated experiments with artificial intelligence (AI) that decides the next set of experiments. Taken to their ultimate expression, SDLs could usher a new paradigm of scientific research, where the world is probed, interpreted, and explained by machines for human benefit. While there are functioning SDLs in the fields of chemistry and materials science, we contend that synthetic biology provides a unique opportunity since the genome provides a single target for affecting the incredibly wide repertoire of biological cell behavior. However, the level of investment required for the creation of biological SDLs is only warranted if directed towards solving difficult and enabling biological questions. Here, we discuss challenges and opportunities in creating SDLs for synthetic biology., Comment: 17 pages, 3 figures. Submitted for publication in Current Opinion in Biotechnology. Updated figure 3 in this version

Sparse regression is frequently employed in diverse scientific settings as a feature selection method. A pervasive aspect of scientific data that hampers both feature selection and estimation is the presence of strong correlations between predictive features. These fundamental issues are often not appreciated by practitioners, and jeapordize conclusions drawn from estimated models. On the other hand, theoretical results on sparsity-inducing regularized regression such as the Lasso have largely addressed conditions for selection consistency via asymptotics, and disregard the problem of model selection, whereby regularization parameters are chosen. In this numerical study, we address these issues through exhaustive characterization of the performance of several regression estimators, coupled with a range of model selection strategies. These estimators and selection criteria were examined across correlated regression problems with varying degrees of signal to noise, distribution of the non-zero model coefficients, and model sparsity. Our results reveal a fundamental tradeoff between false positive and false negative control in all regression estimators and model selection criteria examined. Additionally, we are able to numerically explore a transition point modulated by the signal-to-noise ratio and spectral properties of the design covariance matrix at which the selection accuracy of all considered algorithms degrades. Overall, we find that SCAD coupled with BIC or empirical Bayes model selection performs the best feature selection across the regression problems considered.

Neural optoelectrodes can read and manipulate large numbers of neurons in vivo. However, state-of-the-art devices rely on either standard microfabrication materials (i.e., silicon and silicon nitride), which result in high scalability and throughput but cause severe brain damage due to implant stiffness, or polymeric devices, which are more compliant but whose scalability and implantation in the brain are challenging. Here, we merge the gap between silicon-based fabrication scalability and low (polymeric-like) stiffness by fabricating a nitride and oxide-based optoelectrode with a high density of sensing microelectrodes, passive photonic circuits, and a very small tip thickness (5 μm). We achieve this by removing all the silicon supporting material underneath the probe’s tip—while leaving only the nitride and glass optical ultrathin layers—through a single isotropic etch step. Our optoelectrode integrates 64 electrodes and multiple passive optical outputs, resulting in a cross-sectional area coefficient (the cross section divided by the number of sensors and light emitters) of 3.1—smaller than other optoelectrodes. It also combines a low bending stiffness (∼4.4 × 10−11 N m2), comparable or approaching several state-of-the-art polymeric optoelectrodes. We tested several mechanical insertions of our devices in vivo in rats and demonstrated that we can pierce the pia without using additional temporary supports.

Intracranial electroencephalography (iEEG) data offer a unique combination of high spatial and temporal resolution measures of the living human brain. However, data collection is limited to highly specialized clinical environments. To improve internal (re)use and external sharing of these unique data, we present a structure for storing and sharing iEEG data: BIDS-iEEG, an extension of the Brain Imaging Data Structure (BIDS) specification, along with freely available examples and a bids-starter-kit. BIDS is a framework for organizing and documenting data and metadata with the aim to make datasets more transparent and reusable and to improve reproducibility of research. It is a community-driven specification with an inclusive decision-making process. As an extension of the BIDS specification, BIDS-iEEG facilitates integration with other modalities such as fMRI, MEG, and EEG. As the BIDS-iEEG extension has received input from many iEEG researchers, it provides a common ground for data transfer within labs, between labs, and in open-data repositories. It will facilitate reproducible analyses across datasets, experiments, and recording sites, allowing scientists to answer more complex questions about the human brain. Finally, the cross-modal nature of BIDS will enable efficient consolidation of data from multiple sites for addressing questions about generalized brain function.