Biology subjects

Petrov, D.

Publications and source records attributed to Petrov, D..

9 recordsLinked to original sources

bioRxiv2018

Deep Learning for Quality Control of Subcortical Brain 3D Shape Models

We present several deep learning models for assessing the morphometric fidelity of deep grey matter region models extracted from brain MRI. We test three different convolutional neural net architectures (VGGNet, ResNet and Inception) over 2D maps of geometric features. Further, we present a novel geometry feature augmentation technique based on parametric spherical mapping. Finally, we present an approach for model decision visualization, allowing human raters to see the areas of subcortical shapes most likely to be deemed of failing quality by the machine. Our training data is comprised of 5200 subjects from the ENIGMA Schizophrenia MRI cohorts, and our test dataset contains 1500 subjects from the ENIGMA Major Depressive Disorder cohorts. Our final models reduce human rater time by 46-70%. ResNet outperforms VGGNet and Inception for all of our predictive tasks.

neuroscience↗

bioRxiv2018

Empowering conservation practice with efficient and economical genotyping from poor quality samples

O_LIModerate to high density genotyping (100+ SNPs) is widely used to determine and measure individual identity, relatedness, fitness, population structure and migration in wild populations. C_LIO_LIHowever, these important tools are difficult to apply when high-quality genetic material is unavailable. Most genomic tools are developed for high quality DNA sources from labor medical settings. As a result, most genetic data from market or field settings is limited to easily amplified mitochondrial DNA or a few microsatellites. C_LIO_LITo enable genotyping in conservation contexts, we used next-generation sequencing of multiplex PCR products from very low-quality DNA extracted from feces, hair, and cooked samples. We demonstrated utility and wide-ranging potential application in endangered wild tigers and tracking commercial trade in Caribbean queen conch. C_LIO_LIWe genotyped 100 SNPs from degraded tiger samples to identify individuals, discern close relatives, and detect population differentiation. Co-occurring carnivores do not amplify (e.g. Indian wild dog/Dhole) or are monomorphic (e.g. leopard). 62 SNPs from conch fritters and field-collected samples were used to test relatedness and detect population structure. C_LIO_LIWe provide proof-of-concept for a rapid, simple, cost-effective, and scalable method (for both samples and number of loci), a framework that can be applied to other conservation scenarios previously limited by low quality DNA samples. These approaches provide a critical advance for wildlife monitoring and forensics, open the door to field-ready testing, and will strengthen the use of science in policy decisions and wildlife trade. C_LI

genomics↗

bioRxiv2018

Tissue-specific cis-regulatory divergence implicates a fatty acid elongase necessary for inhibiting interspecies mating in Drosophila

Pheromones known as cuticular hydrocarbons are a major component of reproductive isolation in Drosophila. Individuals from morphologically similar sister species produce different sets of hydrocarbons that allow potential mates to identify them as a suitable partner. In order to explore the molecular mechanisms underlying speciation, we performed RNA-seq in F1 hybrids to measure tissue-specific cis-regulatory divergence between the sister species D. simulans and D. sechellia. By focusing on cis-regulatory changes specific to female oenocytes, we rapidly identified a small number of candidate genes. We found that one of these, the fatty acid elongase eloF, broadly affects both the complement of hydrocarbons present on D. sechellia females and the propensity of D. simulans males to mate with those females. In addition, knockdown of eloF in the more distantly related D. melanogaster led to a similar shift in hydrocarbons as well as lower interspecific mate discrimination by D. simulans males. Thus, cis-regulatory changes in eloF appear to be a major driver in the sexual isolation of D. simulans from multiple other species. More generally, our RNA-seq approach proved to be far more efficient than QTL mapping in identifying candidate genes; the same framework can be used to pinpoint cis-regulatory drivers of divergence in a wide range of traits differing between any interfertile species.

systems biology↗

bioRxiv2018

Broad geographic sampling reveals predictable and pervasive seasonal adaptation in Drosophila

To advance our understanding of adaptation to temporally varying selection pressures, we identified signatures of seasonal adaptation occurring in parallel among Drosophila melanogaster populations. Specifically, we estimated allele frequencies genome-wide from flies sampled early and late in the growing season from 20 widely dispersed populations. We identified parallel seasonal allele frequency shifts across North America and Europe, demonstrating that seasonal adaptation is a general phenomenon of temperate fly populations. Seasonally fluctuating polymorphisms are enriched in large chromosomal inversions and we find a broad concordance between seasonal and spatial allele frequency change. The direction of allele frequency change at seasonally variable polymorphisms can be predicted by weather conditions in the weeks prior to sampling, linking the environment and the genomic response to selection. Our results suggest that fluctuating selection is an important evolutionary force affecting patterns of genetic variation in Drosophila.

evolutionary biology↗

bioRxiv2018

Optically-Controlled \"Living Electrodes\" with Long-Projecting Axon Tracts for a Synaptic Brain-Machine Interface

Achievements in intracortical neural interfaces are compromised by limitations in specificity and long-term performance. A biological intermediary between devices and the brain may offer improved specificity and longevity through natural synaptic integration with deep neural circuitry, while being accessible on the brain surface for optical read-out/control. Accordingly, we have developed the first "living electrodes" comprised of implantable axonal tracts protected within soft hydrogel cylinders for the biologically-mediated monitoring/modulation of brain activity. Here we demonstrate the controlled fabrication, rapid axonal outgrowth, reproducible cytoarchitecture, and simultaneous optical stimulation and recording of neuronal activity within these engineered constructs in vitro. We also present their transplantation, survival, integration, and optical recording in rat cortex in vivo as a proof-of-concept for this neural interface paradigm. The creation and functional validation of these preformed, axon-based "living electrodes" is a critical step towards developing a new class of biohybrid neural interfaces to probe and modulate native circuitry.

bioengineering↗

bioRxiv2018

High accuracy haplotype-derived allele frequencies from ultra-low coverage pool-seq samples

Evolve-and-resequence (E+R) experiments leverage next-generation sequencing technology to track the allele frequency dynamics of populations as they evolve. While previous work has shown that adaptive alleles can be detected by comparing frequency trajectories from many replicate populations, this power comes at the expense of high-coverage (>100x) sequencing of many pooled samples, which can be cost-prohibitive. Here, we show that accurate estimates of allele frequencies can be achieved with very shallow sequencing depths (<5x) via inference of known founder haplotypes in small genomic windows. This technique can be used to efficiently estimate frequencies for any number of bi-allelic SNPs in populations of any model organism founded with sequenced homozygous strains. Using both experimentally-pooled and simulated samples of Drosophila melanogaster, we show that haplotype inference can improve allele frequency accuracy by orders of magnitude for up to 50 generations of recombination, and is robust to moderate levels of missing data, as well as different selection regimes. Finally, we show that a simple linear model generated from these simulations can predict the accuracy of haplotype-derived allele frequencies in other model organisms and experimental designs. To make these results broadly accessible for use in E+R experiments, we introduce HAF-pipe, an open-source software tool for calculating haplotype-derived allele frequencies from raw sequencing data. Ultimately, by reducing sequencing costs without sacrificing accuracy, our method facilitates E+R designs with higher replication and resolution, and thereby, increased power to detect adaptive alleles.

evolutionary biology↗

bioRxiv2017

Entering the era of conservation genomics: Cost-effective assembly of the African wild dog genome using linked long reads

A high-quality reference genome assembly is a valuable tool for the study of non- model organisms across disciplines. Genomic techniques can provide important insights about past population sizes, local adaptation, and even aid in the development of breeding management plans. This information can be particularly important for fields like conservation genetics, where endangered species require critical and immediate attention. However, funding for genomic-based methods can be sparse for conservation projects, as costs for general species management can consume budgets. Here we report the generation of high-quality reference genomes for the African wild dog (Lycaon pictus) at a low cost, thereby facilitating future studies of this endangered canid. We generated assemblies for three individuals from whole blood samples using the linked-read 10x Genomics Chromium system. The most continuous assembly had a scaffold N50 of 21 Mb, a contig N50 of 83 Kb, and completely reconstructed 95% of conserved mammalian genes as reported by BUSCO v2, indicating a high assembly quality. Thus, we show that 10x Genomics Chromium data can be used to effectively generate high-quality genomes of mammal species from Illumina short-read data of intermediate coverage ([~]25-50x). Interestingly, the African wild dog shows a much higher heterozygosity than other species of conservation concern, possibly as a result of its behavioral ecology. The availability of reference genomes for non-model organisms will facilitate better genetic monitoring of threatened species such as the African wild dog. At the same time, they can help researchers and conservationists to better understand the ecology and adaptability of those species in a changing environment.

genomics↗

bioRxiv2016

Deep sequencing of natural and experimental populations of Drosophila melanogaster reveals biases in the spectrum of new mutations

Mutations provide the raw material of evolution, and thus our ability to study evolution depends fundamentally on whether we have precise measurements of mutational rates and patterns. Here we explore the rates and patterns of mutations using i) de novo mutations from Drosophila melanogaster mutation accumulation lines and ii) polymorphisms segregating at extremely low frequencies. The first, mutation accumulation (MA) lines, are the product of maintaining flies in tiny populations for many generations, therefore rendering natural selection ineffective and allowing new mutations to accrue in the genome. In addition to generating a novel dataset of sequenced MA lines, we perform a meta-analysis of all published MA studies in D. melanogaster, which allows more precise estimates of mutational patterns across the genome. In the second half of this work, we identify polymorphisms segregating at extremely low frequencies using several publicly available population genomic data sets from natural populations of D. melanogaster. Extremely rare polymorphisms are difficult to detect with high confidence due to the problem of distinguishing them from sequencing error, however a dataset of true rare polymorphisms would allow the quantification of mutational patterns. This is due to the fact that rare polymorphisms, much like de novo mutations, are on average younger and also relatively unaffected by the filter of natural selection. We identify a high quality set of ~70,000 rare polymorphisms, fully validated with resequencing, and use this dataset to measure mutational patterns in the genome. This includes identifying a high rate of multi-nucleotide mutation events at both short (~5bp) and long (~1kb) genomic distances, showing that mutation drives GC content lower in already GC-poor regions, and finding that the context-dependency of the mutation spectrum predicts long-term evolutionary patterns at four-fold synonymous sites. We also show that de novo mutations from independent mutation accumulation experiments display similar patterns of single nucleotide mutation, and match well the patterns of mutation found in natural populations.

evolutionary biology↗

bioRxiv2016

Pervasive adaptation in Plasmodium-interacting proteins in mammals

The protozoan genus Plasmodium causes malaria in dozens of mammal species, including humans, non-human primates, rodents, and bats. In humans, Plasmodium infections have caused hundreds of millions of documented deaths, imposing strong selection on certain populations and driving the emergence of several resistance alleles. Over the deep timescale of mammalian evolution, however, little is known about host adaptation to Plasmodium. In this work, we expand the collection of known Plasmodium-interacting-proteins (PIPs) in mammalian hosts from ~10 to 410, by manually curating thousands of scientific abstracts. We use comparative tests of adaptation to show that PIPs have experienced >3 times more positive selection than similar mammalian proteins, consistent with Plasmodium as a major and long-standing selective pressure. PIP adaptation is strongly linked to gene expression in the blood, liver, and lung, all of which are clinically relevant tissues in Plasmodium infection. Interestingly, we find that PIPs with immune functions are especially enriched for additional interactions with viruses or bacteria, which together drive a 3.7-fold excess of adaptation. These pleiotropic interactions with unrelated pathogens, along with pressure from other Plasmodium-like Apicomplexan parasites, may help explain the PIP adaptation we observe in all clades of the mammalian tree. As a case study, we also show that alpha-spectrin, the major membrane component of mammalian red blood cells, has experienced accelerated adaptation in domains known to interact specifically with Plasmodium proteins. Similar interactions with Plasmodium-like parasites appear to have driven substantial adaptation in hundreds of host proteins throughout mammalian evolution.

evolutionary biology↗