Deep Neural Networks in Computational Neuroscience, bioRxiv, 2017-05-05
SummaryThe goal of computational neuroscience is to find mechanistic explanations of how the nervous system processes information to give rise to cognitive function and behaviour. At the heart of the field are its models, i.e. mathematical and computational descriptions of the system being studied, which map sensory stimuli to neural responses andor neural to behavioural responses. These models range from simple to complex. Recently, deep neural networks (DNNs) have come to dominate several domains of artificial intelligence (AI). As the term “neural network” suggests, these models are inspired by biological brains. However, current DNNs neglect many details of biological neural networks. These simplifications contribute to their computational efficiency, enabling them to perform complex feats of intelligence, ranging from perceptual (e.g. visual object and auditory speech recognition) to cognitive tasks (e.g. machine translation), and on to motor control (e.g. playing computer games or controlling a robot arm). In addition to their ability to model complex intelligent behaviours, DNNs excel at predicting neural responses to novel sensory stimuli with accuracies well beyond any other currently available model type. DNNs can have millions of parameters, which are required to capture the domain knowledge needed for successful task performance. Contrary to the intuition that this renders them into impenetrable black boxes, the computational properties of the network units are the result of four directly manipulable elements input statistics, network structure, functional objective, and learning algorithm. With full access to the activity and connectivity of all units, advanced visualization techniques, and analytic tools to map network representations to neural data, DNNs represent a powerful framework for building task-performing models and will drive substantial insights in computational neuroscience.
biorxiv neuroscience 100-200-users 2017Gene annotation bias impedes biomedical research, bioRxiv, 2017-05-03
1AbstractWe found tremendous inequality across gene and protein annotation resources. We observe that this bias leads biomedical researchers to focus on richly annotated genes instead of those with the strongest molecular data. We advocate for researchers to reduce these biases by pursuing data-driven hypotheses.
biorxiv bioinformatics 100-200-users 2017Mitochondria are physiologically maintained at close to 50 °C, bioRxiv, 2017-05-03
AbstractIn endothermic species, heat released as a product of metabolism ensures stable internal temperature throughout the organism, despite varying environmental conditions. Mitochondria are major actors in this thermogenic process. Part of the energy released by the oxidation of respiratory substrates drives ATP synthesis and metabolite transport, while a noticeable proportion is released as heat. Using a temperature-sensitive fluorescent probe targeted to mitochondria, we measured mitochondrial temperature in situ under different physiological conditions. At a constant external temperature of 38 °C, mitochondria were more than 10 °C warmer when the respiratory chain was fully functional, both in HEK293cells and primary skin fibroblasts. This differential was abolished in cells lacking mitochondrial DNA or by respiratory inhibitors, but preserved or enhanced by expressing thermogenic enzymes such as the alternative oxidase or the uncoupling protein 1. The activity of various RC enzymes was maximal at, or slightly above, 50 °C. Our study prompts a re-examination of the literature on mitochondria, taking account of the inferred high temperature.
biorxiv cell-biology 200-500-users 2017Reading canonical and modified nucleotides in 16S ribosomal RNA using nanopore direct RNA sequencing, bioRxiv, 2017-04-30
The ribosome small subunit is expressed in all living cells. It performs numerous essential functions during translation, including formation of the initiation complex and proofreading of base-pairs between mRNA codons and tRNA anticodons. The core constituent of the small ribosomal subunit is a ∼1.5 kb RNA strand in prokaryotes (16S rRNA) and a homologous ∼1.8 kb RNA strand in eukaryotes (18S rRNA). Traditional sequencing-by-synthesis (SBS) of rRNA genes or rRNA cDNA copies has achieved wide use as a ‘molecular chronometer’ for phylogenetic studies 1, and as a tool for identifying infectious organisms in the clinic 2. However, epigenetic modifications on rRNA are erased by SBS methods. Here we describe direct MinION nanopore sequencing of individual, full-length 16S rRNA absent reverse transcription or amplification. As little as 5 picograms (∼10 attomole) of E. coli 16S rRNA was detected in 4.5 micrograms of total human RNA. Nanopore ionic current traces that deviated from canonical patterns revealed conserved 16S rRNA base modifications, and a 7-methylguanosine modification that confers aminoglycoside resistance to some pathological E. coli strains. This direct RNA sequencing technology has promise for rapid identification of microbes in the environment and in patient samples.
biorxiv bioengineering 200-500-users 2017Reconstructing the Gigabase Plant Genome of Solanum pennellii using Nanopore Sequencing, bioRxiv, 2017-04-22
Recent updates in sequencing technology have made it possible to obtain Gigabases of sequence data from one single flowcell. Prior to this update, the nanopore sequencing technology was mainly used to analyze and assemble microbial samples1-3. Here, we describe the generation of a comprehensive nanopore sequencing dataset with a median fragment size of 11,979 bp for the wild tomato species Solanum pennellii featuring an estimated genome size of ca 1.0 to 1.1 Gbases. We describe its genome assembly to a contig N50 of 2.5 MB using a pipeline comprising a Canu4 pre-processing and a subsequent assembly using SMARTdenovo. We show that the obtained nanopore based de novo genome reconstruction is structurally highly similar to that of the reference S. pennellii LA7165 genome but has a high error rate caused mostly by deletions in homopolymers. After polishing the assembly with Illumina short read data we obtained an error rate of <0.02 % when assessed versus the same Illumina data. More importantly however we obtained a gene completeness of 96.53% which even slightly surpasses that of the reference S. pennellii genome5. Taken together our data indicate such long read sequencing data can be used to affordably sequence and assemble Gbase sized diploid plant genomes.Raw data is available at <jatsext-link xmlnsxlink=httpwww.w3.org1999xlink ext-link-type=uri xlinkhref=httpwww.plabipd.deportalsolanum-pennellii>httpwww.plabipd.deportalsolanum-pennellii<jatsext-link> and has been deposited as PRJEB19787.
biorxiv genomics 0-100-users 2017Nanopore sequencing and assembly of a human genome with ultra-long reads, bioRxiv, 2017-04-21
AbstractNanopore sequencing is a promising technique for genome sequencing due to its portability, ability to sequence long reads from single molecules, and to simultaneously assay DNA methylation. However until recently nanopore sequencing has been mainly applied to small genomes, due to the limited output attainable. We present nanopore sequencing and assembly of the GM12878 UtahCeph human reference genome generated using the Oxford Nanopore MinION and R9.4 version chemistry. We generated 91.2 Gb of sequence data (∼30× theoretical coverage) from 39 flowcells. De novo assembly yielded a highly complete and contiguous assembly (NG50 ∼3Mb). We observed considerable variability in homopolymeric tract resolution between different basecallers. The data permitted sensitive detection of both large structural variants and epigenetic modifications. Further we developed a new approach exploiting the long-read capability of this system and found that adding an additional 5×-coverage of ‘ultra-long’ reads (read N50 of 99.7kb) more than doubled the assembly contiguity. Modelling the repeat structure of the human genome predicts extraordinarily contiguous assemblies may be possible using nanopore reads alone. Portable de novo sequencing of human genomes may be important for rapid point-of-care diagnosis of rare genetic diseases and cancer, and monitoring of cancer progression. The complete dataset including raw signal is available as an Amazon Web Services Open Dataset at <jatsext-link xmlnsxlink=httpwww.w3.org1999xlink ext-link-type=uri xlinkhref=httpsgithub.comnanopore-wgs-consortiumNA12878>httpsgithub.comnanopore-wgs-consortiumNA12878<jatsext-link>.
biorxiv genomics 500+-users 2017