A community-powered search of machine learning strategy space to find NMR property prediction models.

The rise of machine learning (ML) has created an explosion in the potential strategies for using data to make scientific predictions. For physical scientists wishing to apply ML strategies to a particular domain, it can be difficult to assess in advance what strategy to adopt within a vast space of possibilities. Here we outline the results of an online community-powered effort to swarm search the space of ML strategies and develop algorithms for predicting atomic-pairwise nuclear magnetic resonance (NMR) properties in molecules. Using an open-source dataset, we worked with Kaggle to design and host a 3-month competition which received 47,800 ML model predictions from 2,700 teams in 84 countries. Within 3 weeks, the Kaggle community produced models with comparable accuracy to our best previously published 'in-house' efforts. A meta-ensemble model constructed as a linear combination of the top predictions has a prediction accuracy which exceeds that of any individual model, 7-19x better than our previous state-of-the-art. The results highlight the potential of transformer architectures for predicting quantum mechanical (QM) molecular properties.

Capacitive Heart-Rate Sensing on Touch Screen Panel with Laterally Interspaced Electrodes.

It is demonstrated that the heart-rate can be sensed capacitively on a touch screen panel (TSP) together with touch signals. The existing heart-rate sensing systems measure blood pulses by tracing the amount of light reflected from blood vessels during a number of cardiac cycles. This type of sensing system requires a considerable amount of power and space to be implemented in multi-functional mobile devices such as smart phones. It is found that the variation of the effective dielectric constant of finger stemming from the difference of systolic and diastolic blood flows can be measured with laterally interspaced top electrodes of TSP. The spacing between a pair of non-adjacent top electrodes turns out to be wide enough to distinguish heart-rate signals from noises. With the aid of fast Fourier transform, the heart-rate can be extracted reliably, which matches with the one obtained by actually counting heart beats on the wrist.

Negative Fermi-Level Pinning Effect of Metal/n-GaAs(001) Junction Induced by a Graphene Interlayer.

It is demonstrated that the electric dipole layer due to the overlapping of electron wave functions at the metal/graphene contact results in a negative Fermi-level pinning effect on the region of the GaAs surface with low interface-trap density in the metal/graphene/n-GaAs(001) junction. The graphene interlayer plays the role of a diffusion barrier, preventing the atomic intermixing at the interface and preserving the low interface-trap density region. The negative Fermi-level pinning effect is supported by the decrease of the Schottky barrier with the increase of the metal work function. Our work shows that the graphene interlayer can invert the effective work function of the metal between high and low, making it possible to form both Schottky and Ohmic-like contacts with identical (particularly high work function) metal electrodes on a semiconductor substrate possessing low surface-state density.

Complete sequencing and comparative analyses of the pepper (Capsicum annuum L.) plastome revealed high frequency of tandem repeats and large insertion/deletions on pepper plastome.

Plants in the family Solanaceae are used as model systems in comparative and evolutionary genomics. The complete chloroplast genomes of seven solanaceous species have been sequenced, including tobacco, potato and tomato, but not peppers. We analyzed the complete chloroplast genome sequence of the hot pepper, Capsicum annuum. The pepper chloroplast genome was 156,781 bp in length, including a pair of inverted repeats (IR) of 25,783 bp. The content and the order of 133 genes in the pepper chloroplast genome were identical to those of other solanaceous plastomes. To characterize pepper plastome sequence, we performed comparative analysis using complete plastome sequences of pepper and seven solanaceous plastomes. Frequency and contents of large indels and tandem repeat sequences and distribution pattern of genome-wide sequence variations were investigated. In addition, a phylogenetic analysis using concatenated alignments of coding sequences was performed to determine evolutionary position of pepper in Solanaceae. Our results revealed two distinct features of pepper plastome compared to other solanaceous plastomes. Firstly, large indels, including insertions on accD and rpl20 gene sequences, were predominantly detected in the pepper plastome compared to other solanaceous plastomes. Secondly, tandem repeat sequences were particularly frequent in the pepper plastome. Taken together, our study represents unique features of evolution of pepper plastome among solanaceous plastomes.

IMGD: an integrated platform supporting comparative genomics and phylogenetics of insect mitochondrial genomes.

BACKGROUND: Sequences and organization of the mitochondrial genome have been used as markers to investigate evolutionary history and relationships in many taxonomic groups. The rapidly increasing mitochondrial genome sequences from diverse insects provide ample opportunities to explore various global evolutionary questions in the superclass Hexapoda. To adequately support such questions, it is imperative to establish an informatics platform that facilitates the retrieval and utilization of available mitochondrial genome sequence data. RESULTS: The Insect Mitochondrial Genome Database (IMGD) is a new integrated platform that archives the mitochondrial genome sequences from 25,747 hexapod species, including 112 completely sequenced and 20 nearly completed genomes and 113,985 partially sequenced mitochondrial genomes. The Species-driven User Interface (SUI) of IMGD supports data retrieval and diverse analyses at multi-taxon levels. The Phyloviewer implemented in IMGD provides three methods for drawing phylogenetic trees and displays the resulting trees on the web. The SNP database incorporated to IMGD presents the distribution of SNPs and INDELs in the mitochondrial genomes of multiple isolates within eight species. A newly developed comparative SNU Genome Browser supports the graphical presentation and interactive interface for the identified SNPs/INDELs. CONCLUSION: The IMGD provides a solid foundation for the comparative mitochondrial genomics and phylogenetics of insects. All data and functions described here are available at the web site (http://www.imgd.org/).