Unsupervised discovery of temporal sequences in high-dimensional datasets, with applications to neuroscience.

Identifying low-dimensional features that describe large-scale neural recordings is a major challenge in neuroscience. Repeated temporal patterns (sequences) are thought to be a salient feature of neural dynamics, but are not succinctly captured by traditional dimensionality reduction techniques. Here, we describe a software toolbox-called seqNMF-with new methods for extracting informative, non-redundant, sequences from high-dimensional neural data, testing the significance of these extracted patterns, and assessing the prevalence of sequential structure in data. We test these methods on simulated data under multiple noise conditions, and on several real neural and behavioral datas. In hippocampal data, seqNMF identifies neural sequences that match those calculated manually by reference to behavioral events. In songbird data, seqNMF discovers neural sequences in untutored birds that lack stereotyped songs. Thus, by identifying temporal structure directly from neural data, seqNMF enables dissection of complex neural circuits without relying on temporal references from stimuli or behavioral outputs.

Differentially timed extracellular signals synchronize pacemaker neuron clocks.

Synchronized neuronal activity is vital for complex processes like behavior. Circadian pacemaker neurons offer an unusual opportunity to study synchrony as their molecular clocks oscillate in phase over an extended timeframe (24 h). To identify where, when, and how synchronizing signals are perceived, we first studied the minimal clock neural circuit in Drosophila larvae, manipulating either the four master pacemaker neurons (LNvs) or two dorsal clock neurons (DN1s). Unexpectedly, we found that the PDF Receptor (PdfR) is required in both LNvs and DN1s to maintain synchronized LNv clocks. We also found that glutamate is a second synchronizing signal that is released from DN1s and perceived in LNvs via the metabotropic glutamate receptor (mGluRA). Because simultaneously reducing Pdfr and mGluRA expression in LNvs severely dampened Timeless clock protein oscillations, we conclude that the master pacemaker LNvs require extracellular signals to function normally. These two synchronizing signals are released at opposite times of day and drive cAMP oscillations in LNvs. Finally we found that PdfR and mGluRA also help synchronize Timeless oscillations in adult s-LNvs. We propose that differentially timed signals that drive cAMP oscillations and synchronize pacemaker neurons in circadian neural circuits will be conserved across species.