Contact-dependent cell-cell communications drive morphological invariance during ascidian embryogenesis, bioRxiv, 2017-12-25
ABSTRACTCanalization of developmental processes ensures the reproducibility and robustness of embryogenesis within each species. In its extreme form, found in ascidians, early embryonic cell lineages are invariant between embryos within and between species, despite rapid genomic divergence. To resolve this paradox, we used live light-sheet imaging to quantify individual cell behaviors in digitalized embryos and explore the forces that canalize their development. This quantitative approach revealed that individual cell geometries and cell contacts are strongly constrained, and that these constraints are tightly linked to the control of fate specification by local cell inductions. While in vertebrates ligand concentration usually controls cell inductions, we found that this role is fulfilled in ascidians by the area of contacts between signaling and responding cells. We propose that the duality between geometric and genetic control of inductions contributes to the counterintuitive inverse correlation between geometric and genetic variability during embryogenesis.
biorxiv developmental-biology 100-200-users 2017DeepMHC Deep Convolutional Neural Networks for High-performance peptide-MHC Binding Affinity Prediction, bioRxiv, 2017-12-25
AbstractConvolutional neural networks (CNN) have been shown to outperform conventional methods in DNA-protien binding specificity prediction. However, whether we can transfer this success to protien-peptide binding affinity prediction depends on appropriate design of the CNN architectue that calls for thorough understanding how to match the architecture to the problem. Here we propose DeepMHC, a deep convolutional neural network (CNN) based protein-peptide binding prediction algorithm for achieving better performance in MHC-I peptide binding affinity prediction than conventional algorithms. Our model takes only raw binding peptide sequences as input without needing any human-designed features and othe physichochemical or evolutionary information of the amino acids. Our CNN models are shown to be able to learn non-linear relationships among the amino acid positions of the peptides to achieve highly competitive performance on most of the IEDB benchmark datasets with a single model architecture and without using any consensus or composite ensemble classifier models. By systematically exploring the best CNN architecture, we identified critical design considerations in CNN architecture development for peptide-MHC binding prediction.
biorxiv bioinformatics 100-200-users 2017MAPLE a Modular Automated Platform for Large-scale Experiments, a low-cost robot for integrated animal-handling and phenotyping, bioRxiv, 2017-12-25
AbstractGenetic model system animals have significant scientific value in part because of large-scale experiments like screens, but performing such experiments over long time periods by hand is arduous and risks errors. Thus the field is poised to benefit from automation, just as molecular biology did from liquid-handling robots. We developed a Modular Automated Platform for Large-scale Experiments (MAPLE), a Drosophila-handling robot capable of conducting lab tasks and experiments. We demonstrate MAPLE’s ability to accelerate the collection of virgin female flies (a pervasive experimental chore in fly genetics) and assist high-throughput phenotyping assays. Using MAPLE to autonomously run a novel social interaction experiment, we found that 1) pairs of flies exhibit persistent idiosyncrasies in affiliative behavior, 2) these dyad-specific interactions require olfactory and visual cues, and 3) social interaction network structure is topologically stable over time. These diverse examples demonstrate MAPLE’s versatility as a general platform for conducting fly science automatically.
biorxiv bioengineering 0-100-users 2017Discovery and characterization of coding and non-coding driver mutations in more than 2,500 whole cancer genomes, bioRxiv, 2017-12-24
AbstractDiscovery of cancer drivers has traditionally focused on the identification of protein-coding genes. Here we present a comprehensive analysis of putative cancer driver mutations in both protein-coding and non-coding genomic regions across >2,500 whole cancer genomes from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium. We developed a statistically rigorous strategy for combining significance levels from multiple driver discovery methods and demonstrate that the integrated results overcome limitations of individual methods. We combined this strategy with careful filtering and applied it to protein-coding genes, promoters, untranslated regions (UTRs), distal enhancers and non-coding RNAs. These analyses redefine the landscape of non-coding driver mutations in cancer genomes, confirming a few previously reported elements and raising doubts about others, while identifying novel candidate elements across 27 cancer types. Novel recurrent events were found in the promoters or 5’UTRs of TP53, RFTN1, RNF34, and MTG2, in the 3’UTRs of NFKBIZ and TOB1, and in the non-coding RNA RMRP. We provide evidence that the previously reported non-coding RNAs NEAT1 and MALAT1 may be subject to a localized mutational process. Perhaps the most striking finding is the relative paucity of point mutations driving cancer in non-coding genes and regulatory elements. Though we have limited power to discover infrequent non-coding drivers in individual cohorts, combined analysis of promoters of known cancer genes show little excess of mutations beyond TERT.
biorxiv genomics 100-200-users 2017Emergence of reward expectation signals in identified dopamine neurons, bioRxiv, 2017-12-23
AbstractCoherent control of purposive actions emerges from the coordination of multiple brain circuits during learning. Dissociable brain circuits and cell-types are thought to preferentially participate in distinct learning mechanisms. For example, the activity of midbrain dopamine (mDA) neurons is proposed to primarily, or even exclusively, reflect reward prediction error signals in well-trained animals. To study the specific contribution of individual circuits requires observing changes before tight functional coordination is achieved. However, little is known about the detailed timing of the emergence of reward-related representations in dopaminergic neurons. Here we recorded activity of identified dopaminergic neurons as naïve mice learned a novel stimulus-reward association. We found that at early stages of learning mDA neuron activity reflected both external (sensory) and internal (action initiation) causes of reward expectation. The increasingly precise correlation of action initiation with sensory stimuli rather than an evaluation of outcomes governed mDA neuron activity. Thus, our data demonstrate that mDA neuron activity early in learning does not reflect errors, but is more akin to a Hebbian learning signal - providing new insight into a critical computation in a highly conserved, essential learning circuit.
biorxiv neuroscience 100-200-users 2017A single-cell catalogue of regulatory states in the ageing Drosophila brain, bioRxiv, 2017-12-22
SummaryThe diversity of cell types and regulatory states in the brain, and how these change during ageing, remains largely unknown. Here, we present a single-cell transcriptome catalogue of the entire adult Drosophila melanogaster brain sampled across its lifespan. Both neurons and glia age through a process of “regulatory erosion”, characterized by a strong decline of RNA content, and accompanied by increasing transcriptional and chromatin noise. We identify more than 50 cell types by specific transcription factors and their downstream gene regulatory networks. In addition to neurotransmitter types and neuroblast lineages, we find a novel neuronal cell state driven by datilografo and prospero. This state relates to neuronal birth order, the metabolic profile, and the activity of a neuron. Our single-cell brain catalogue reveals extensive regulatory heterogeneity linked to ageing and brain function and will serve as a reference for future studies of genetic variation and disease mutations.
biorxiv genomics 0-100-users 2017