The formation of neuronal connections requires the precise guidance of developing axons toward their targets. In the Drosophila visual system, photoreceptor neurons (R cells) project from the eye into the brain. These cells are grouped into some 750 clusters comprised of eight photoreceptors or R cells each. R cells fall into three classes: R1 to R6, R7, and R8. Posterior R8 cells are the first to project axons into the brain. How these axons select a specific pathway is not known. Here, we used a microarray-based approach to identify genes expressed in R8 neurons as they extend into the brain. We found that Roundabout-3 (Robo3), an axon-guidance receptor, is expressed specifically and transiently in R8 growth cones. In wild-type animals, posterior-most R8 axons extend along a border of glial cells demarcated by the expression of Slit, the secreted ligand of Robo3. In contrast, robo3 mutant R8 axons extend across this border and fasciculate inappropriately with other axon tracts. We demonstrate that either Robo1 or Robo2 rescues the robo3 mutant phenotype when each is knocked into the endogenous robo3 locus separately, indicating that R8 does not require a function unique to the Robo3 paralog. However, persistent expression of Robo3 in R8 disrupts the layer-specific targeting of R8 growth cones. Thus, the transient cell-specific expression of Robo3 plays a crucial role in establishing neural circuits in the Drosophila visual system by selectively regulating pathway choice for posterior-most R8 growth cones.

Electrophysiology is one of the major experimental techniques used in neuroscience. The favorable spatial and temporal resolution as well as the increasingly larger site counts of brain recording electrodes contribute to the popularity and importance of electrophysiology in neuroscience. Such electrodes are typically mechanically placed in the brain to perform acute or chronic freely moving animal measurements. The micro positioners currently used for such tasks employ a single translator per independent probe being placed into the targeted brain region, leading to significant size and weight restrictions. To overcome this limitation, we have developed a miniature robotic multi-probe neural microdrive that utilizes novel phase-change-material-filled resistive heater micro-grippers. The microscopic dimensions, gentle gripping action, independent electronic actuation control, and high packing density of the grippers allow for micrometer-precision independent positioning of multiple arbitrarily shaped parallel neural electrodes with only a single piezo actuator in an inchworm motor configuration. This multi-probe-single-actuator design allows for significant size and weight reduction, as well as remote control and potential automation of the microdrive. We demonstrate accurate placement of multiple independent recording electrodes into the CA1 region of the rat hippocampus in vivo in acute and chronic settings. Thus, our robotic neural microdrive technology is applicable towards basic neuroscience and clinical studies, as well as other multi-probe or multi-sensor micro-positioning applications.

Mammalian development is characterized with transitions from homogeneous populations of precursor to heterogeneous population of specified cells. We review here the main dynamical mechanisms through which such transitions are conceptualized, and discuss that the differentiation timing, robust cell-type proportions and recovery upon perturbation are emergent property of proliferating and communicating cell populations. We argue that studying developmental systems using transitions in collective system states is necessary to describe observed experimental features, and propose additionally the basis of a novel analytical method to deduce the relationship between single-cell dynamics and the collective, symmetry-broken states in cellular populations.

Small central pattern generating circuits found in invertebrates have significant advantages for the study of the circuit mechanisms that generate brain rhythms. Experimental and computational studies of small oscillatory circuits reveal that similar rhythms can arise from disparate mechanisms. Animal-to-animal variation in the properties of single neurons and synapses may underly robust circuit performance, and can be revealed by perturbations. Neuromodulation can produce altered circuit performance but also ensure reliable circuit function.

We consider the problem of estimating discrete selfexciting point process models from limited binary observations, where the history of the process serves as the covariate. We analyze the performance of two classes of estimators, namely the `1-regularized maximum likelihood and greedy estimators, for a canonical self-exciting point process and characterize the sampling tradeoffs required for stable recovery in the non-asymptotic regime. Our results extend those of compressed sensing for linear and generalized linear models with i.i.d. covariates to point processes with highly inter-dependent covariates. We further provide simulation studies as well as application to real spiking data from mouse’s lateral geniculate nucleus and ferret’s retinal ganglion cells which agree with our theoretical predictions.

Single-particle tracking (SPT) has become an important method to bridge biochemistry and cell biology since it allows direct observation of protein binding and diffusion dynamics in live cells. However, accurately inferring information from SPT studies is challenging due to biases in both data analysis and experimental design. To address analysis bias, we introduce 'Spot-On', an intuitive web-interface. Spot-On implements a kinetic modeling framework that accounts for known biases, including molecules moving out-of-focus, and robustly infers diffusion constants and subpopulations from pooled single-molecule trajectories. To minimize inherent experimental biases, we implement and validate stroboscopic photo-activation SPT (spaSPT), which minimizes motion-blur bias and tracking errors. We validate Spot-On using experimentally realistic simulations and show that Spot-On outperforms other methods. We then apply Spot-On to spaSPT data from live mammalian cells spanning a wide range of nuclear dynamics and demonstrate that Spot-On consistently and robustly infers subpopulation fractions and diffusion constants.

Neural activity maintains representations that bridge past and future events, often over many seconds. Network models can produce persistent and ramping activity, but the positive feedback that is critical for these slow dynamics can cause sensitivity to perturbations. Here we use electrophysiology and optogenetic perturbations in the mouse premotor cortex to probe the robustness of persistent neural representations during motor planning. We show that preparatory activity is remarkably robust to large-scale unilateral silencing: detailed neural dynamics that drive specific future movements were quickly and selectively restored by the network. Selectivity did not recover after bilateral silencing of the premotor cortex. Perturbations to one hemisphere are thus corrected by information from the other hemisphere. Corpus callosum bisections demonstrated that premotor cortex hemispheres can maintain preparatory activity independently. Redundancy across selectively coupled modules, as we observed in the premotor cortex, is a hallmark of robust control systems. Network models incorporating these principles show robustness that is consistent with data.

Multi-modal image registration is a challenging task that is vital to fuse complementary signals for subsequent analyses. Despite much research into cost functions addressing this challenge, there exist cases in which these are ineffective. In this work, we show that (1) this is true for the registration of in-vivo Drosophila brain volumes visualizing genetically encoded calcium indicators to an nc82 atlas and (2) that machine learning based contrast synthesis can yield improvements. More specifically, the number of subjects for which the registration outright failed was greatly reduced (from 40% to 15%) by using a synthesized image.

Calcium imaging is a powerful method to record the activity of neural populations, but inferring spike times from calcium signals is a challenging problem. We compared multiple approaches using multiple datasets with ground truth electrophysiology, and found that simple non-negative deconvolution (NND) outperformed all other algorithms. We introduce a novel benchmark applicable to recordings without electrophysiological ground truth, based on the correlation of responses to two stimulus repeats, and used this to show that unconstrained NND also outperformed the other algorithms when run on 'zoomed out' datasets of ~10,000 cell recordings. Finally, we show that NND-based methods match the performance of a supervised method based on convolutional neural networks, while avoiding some of the biases of such methods, and at much faster running times. We therefore recommend that spikes be inferred from calcium traces using simple NND, due to its simplicity, efficiency and accuracy.

Calcium imaging is a powerful method to record the activity of neural populations in many species, but inferring spike times from calcium signals is a challenging problem. We compared multiple approaches using multiple datasets with ground truth electrophysiology, and found that simple non-negative deconvolution (NND) outperformed all other algorithms on out-of-sample test data. We introduce a novel benchmark applicable to recordings without electrophysiological ground truth, based on the correlation of responses to two stimulus repeats, and used this to show that unconstrained NND also outperformed the other algorithms when run on "zoomed out" datasets of ∼10,000 cell recordings from the visual cortex of mice of either sex. Finally, we show that NND-based methods match the performance of a supervised method based on convolutional neural networks, while avoiding some of the biases of such methods, and at much faster running times. We therefore recommend that spikes be inferred from calcium traces using simple NND, due to its simplicity, efficiency and accuracy.The experimental method that currently allows for recordings of the largest numbers of cells simultaneously is two-photon calcium imaging. However, use of this powerful method requires that neuronal firing times be inferred correctly from the large resulting datasets. Previous studies have claimed that complex supervised learning algorithms outperform simple deconvolution methods at this task. Unfortunately, these studies suffered from several problems and biases. When we repeated the analysis, using the same data and correcting these problems, we found that simpler spike inference methods perform better. Even more importantly, we found that supervised learning methods can introduce artifactual structure into spike trains, that can in turn lead to erroneous scientific conclusions. Of the algorithms we evaluated, we found that an extremely simple method performed best in all circumstances tested, was much faster to run, and was insensitive to parameter choices, making incorrect scientific conclusions much less likely.