Fast, sensitive, and accurate integration of single cell data with Harmony, bioRxiv, 2018-11-05
AbstractThe rapidly emerging diversity of single cell RNAseq datasets allows us to characterize the transcriptional behavior of cell types across a wide variety of biological and clinical conditions. With this comprehensive breadth comes a major analytical challenge. The same cell type across tissues, from different donors, or in different disease states, may appear to express different genes. A joint analysis of multiple datasets requires the integration of cells across diverse conditions. This is particularly challenging when datasets are assayed with different technologies in which real biological differences are interspersed with technical differences. We present Harmony, an algorithm that projects cells into a shared embedding in which cells group by cell type rather than dataset-specific conditions. Unlike available single-cell integration methods, Harmony can simultaneously account for multiple experimental and biological factors. We develop objective metrics to evaluate the quality of data integration. In four separate analyses, we demonstrate the superior performance of Harmony to four single-cell-specific integration algorithms. Moreover, we show that Harmony requires dramatically fewer computational resources. It is the only available algorithm that makes the integration of ∼ 106 cells feasible on a personal computer. We demonstrate that Harmony identifies both broad populations and fine-grained subpopulations of PBMCs from datasets with large experimental differences. In a meta-analysis of 14,746 cells from 5 studies of human pancreatic islet cells, Harmony accounts for variation among technologies and donors to successfully align several rare subpopulations. In the resulting integrated embedding, we identify a previously unidentified population of potentially dysfunctional alpha islet cells, enriched for genes active in the Endoplasmic Reticulum (ER) stress response. The abundance of these alpha cells correlates across donors with the proportion of dysfunctional beta cells also enriched in ER stress response genes. Harmony is a fast and flexible general purpose integration algorithm that enables the identification of shared fine-grained subpopulations across a variety of experimental and biological conditions.
biorxiv bioinformatics 100-200-users 2018Neural Population Control via Deep Image Synthesis, bioRxiv, 2018-11-05
Particular deep artificial neural networks (ANNs) are today’s most accurate models of the primate brain’s ventral visual stream. Here we report that, using a targeted ANN-driven image synthesis method, new luminous power patterns (i.e. images) can be applied to the primate retinae to predictably push the spiking activity of targeted V4 neural sites beyond naturally occurring levels. More importantly, this method, while not yet perfect, already achieves unprecedented independent control of the activity state of entire populations of V4 neural sites, even those with overlapping receptive fields. These results show how the knowledge embedded in today’s ANN models might be used to non-invasively set desired internal brain states at neuron-level resolution, and suggest that more accurate ANN models would produce even more accurate control.
biorxiv neuroscience 100-200-users 2018On the adaptive behavior of head-fixed flies navigating in two-dimensional, visual virtual reality, bioRxiv, 2018-11-05
AbstractA navigating animal’s sensory experience is shaped not just by its surroundings, but by its movements within them, which in turn are influenced by its past experiences. Studying the intertwined roles of sensation, experience and directed action in navigation has been made easier by the development of virtual reality (VR) environments for head-fixed animals, which allow for quantitative measurements of behavior in well-controlled sensory conditions. VR has long featured in studies of Drosophila melanogaster, but these experiments have typically relied on one-dimensional (1D) VR, effectively allowing the fly to change only its heading in a visual scene, and not its position. Here we explore how flies navigate in a two-dimensional (2D) visual VR environment that more closely resembles their experience during free behavior. We show that flies’ interaction with landmarks in 2D environments cannot be automatically derived from their behavior in simpler 1D environments. Using a novel paradigm, we then demonstrate that flies in 2D VR adapt their behavior in a visual environment in response to optogenetically delivered appetitive and aversive stimuli. Much like free-walking flies after encounters with food, head-fixed flies respond to optogenetic activation of sugar-sensing neurons by initiating a local search behavior. Finally, by pairing optogenetic activation of heat-sensing cells to the flies’ presence near visual landmarks of specific shapes, we elicit selective learned avoidance of landmarks associated with aversive “virtual heat”. These head-fixed paradigms set the stage for an interrogation of fly brain circuitry underlying flexible navigation in complex visual environments.
biorxiv animal-behavior-and-cognition 0-100-users 2018tartan underlies the evolution of male Drosophila genital morphology, bioRxiv, 2018-11-05
AbstractMale genital structures are among the most rapidly evolving morphological traits and are often the only features that can distinguish closely related species. This process is thought to be driven by sexual selection and may reinforce species separation. However, while the genetic basis of many phenotypic differences have been identified, we still lack knowledge about the genes underlying evolutionary differences in male genital organs and organ size more generally. The claspers (surstyli) are periphallic structures that play an important role in copulation in insects. Here we show that natural variation in clasper size and bristle number between Drosophila mauritiana and D. simulans is caused by evolutionary changes in tartan (trn), which encodes a transmembrane leucine-rich repeat domain protein that mediates cell-cell interactions and affinity differences. There are no fixed amino acid differences in trn between D. mauritiana and D. simulans but differences in the expression of this gene in developing genitalia suggest cis-regulatory changes in trn underlie the evolution of clasper morphology in these species. Finally, analysis of reciprocal hemizyotes that are genetically identical, except for which species the functional allele of trn is from, determined that the trn allele of D. mauritiana specifies larger claspers with more bristles than the allele of D. simulans. Therefore we have identified the first gene underlying evolutionary change in the size of a male genital organ, which will help to better understand the rapid diversification of these structures and the regulation and evolution of organ size more broadly.Significance StatementThe morphology of male genital organs evolves rapidly driven by sexual selection. However, little is known about the genes underlying genitalia differences between species. Identifying these genes is key to understanding how sexual selection acts on development to produce rapid phenotypic change. We have found that the gene tartan underlies differences between male Drosophila mauritiana and D. simulans in the size and bristle number of the claspers - genital projections that grasp the female during copulation. Moreover, since tartan encodes a protein that is involved in cell affinity, this may represent a new developmental mechanism for morphological change. Therefore, our study provides new insights into genetic and developmental bases for the rapid evolution of male genitalia and organ size more generally.
biorxiv evolutionary-biology 0-100-users 2018Valid post-clustering differential analysis for single-cell RNA-Seq, bioRxiv, 2018-11-05
SummarySingle-cell computational pipelines involve two critical steps organizing cells (clustering) and identifying the markers driving this organization (differential expression analysis). State-of-the-art pipelines perform differential analysis after clustering on the same dataset. We observe that because clustering forces separation, reusing the same dataset generates artificially low p-values and hence false discoveries. We introduce a valid post-clustering differential analysis framework which corrects for this problem. We provide software at <jatsext-link xmlnsxlink=httpwww.w3.org1999xlink ext-link-type=uri xlinkhref=httpsgithub.comjessemzhangtn_test>httpsgithub.comjessemzhangtn_test<jatsext-link>.
biorxiv bioinformatics 100-200-users 2018Systematic identification of human SNPs affecting regulatory element activity, bioRxiv, 2018-11-04
AbstractMost of the millions of single-nucleotide polymorphisms (SNPs) in the human genome are non-coding, and many overlap with putative regulatory elements. Genome-wide association studies have linked many of these SNPs to human traits or to gene expression levels, but rarely with sufficient resolution to identify the causal SNPs. Functional screens based on reporter assays have previously been of insufficient throughput to test the vast space of SNPs for possible effects on enhancer and promoter activity. Here, we have leveraged the throughput of the SuRE reporter technology to survey a total of 5.9 million SNPs, including 57% of the known common SNPs. We identified more than 30 thousand SNPs that alter the activity of putative regulatory elements, often in a cell-type specific manner. These data indicate that a large proportion of human non-coding SNPs may affect gene regulation. Integration of these SuRE data with genome-wide association studies may help pinpoint SNPs that underlie human traits.
biorxiv genomics 100-200-users 2018