Genomic sketching with multiplicities and locality-sensitive hashing using Dashing 2.

A genomic sketch is a small, probabilistic representation of the set of k-mers in a sequencing data set. Sketches are building blocks for large-scale analyses that consider similarities between many pairs of sequences or sequence collections. Although existing tools can easily compare tens of thousands of genomes, data sets can reach millions of sequences and beyond. Popular tools also fail to consider k-mer multiplicities, making them less applicable in quantitative settings. Here, we describe a method called Dashing 2 that builds on the SetSketch data structure. SetSketch is related to HyperLogLog (HLL) but discards use of leading zero count in favor of a truncated logarithm of adjustable base. Unlike HLL, SetSketch can perform multiplicity-aware sketching when combined with the ProbMinHash method. Dashing 2 integrates locality-sensitive hashing to scale all-pairs comparisons to millions of sequences. It achieves superior similarity estimates for the Jaccard coefficient and average nucleotide identity compared with the original Dashing, but in much less time while using the same-sized sketch. Dashing 2 is a free, open source software.

A scalable and unbiased discordance metric with H.

A standard unsupervised analysis is to cluster observations into discrete groups using a dissimilarity measure, such as Euclidean distance. If there does not exist a ground-truth label for each observation necessary for external validity metrics, then internal validity metrics, such as the tightness or separation of the clusters, are often used. However, the interpretation of these internal metrics can be problematic when using different dissimilarity measures as they have different magnitudes and ranges of values that they span. To address this problem, previous work introduced the "scale-agnostic" $G_{+}$ discordance metric; however, this internal metric is slow to calculate for large data. Furthermore, in the setting of unsupervised clustering with $k$ groups, we show that $G_{+}$ varies as a function of the proportion of observations assigned to each of the groups (or clusters), referred to as the group balance, which is an undesirable property. To address this problem, we propose a modification of $G_{+}$, referred to as $H_{+}$, and demonstrate that $H_{+}$ does not vary as a function of group balance using a simulation study and with public single-cell RNA-sequencing data. Finally, we provide scalable approaches to estimate $H_{+}$, which are available in the $\mathtt{fasthplus}$ R package.

Fast and memory-efficient scRNA-seq k-means clustering with various distances.

Single-cell RNA-sequencing (scRNA-seq) analyses typically begin by clustering a gene-by-cell expression matrix to empirically define groups of cells with similar expression profiles. We describe new methods and a new open source library, minicore, for efficient k-means++ center finding and k-means clustering of scRNA-seq data. Minicore works with sparse count data, as it emerges from typical scRNA-seq experiments, as well as with dense data from after dimensionality reduction. Minicore's novel vectorized weighted reservoir sampling algorithm allows it to find initial k-means++ centers for a 4-million cell dataset in 1.5 minutes using 20 threads. Minicore can cluster using Euclidean distance, but also supports a wider class of measures like Jensen-Shannon Divergence, Kullback-Leibler Divergence, and the Bhattachaiyya distance, which can be directly applied to count data and probability distributions. Further, minicore produces lower-cost centerings more efficiently than scikit-learn for scRNA-seq datasets with millions of cells. With careful handling of priors, minicore implements these distance measures with only minor (<2-fold) speed differences among all distances. We show that a minicore pipeline consisting of k-means++, localsearch++ and mini-batch k-means can cluster a 4-million cell dataset in minutes, using less than 10GiB of RAM. This memory-efficiency enables atlas-scale clustering on laptops and other commodity hardware. Finally, we report findings on which distance measures give clusterings that are most consistent with known cell type labels.

Wave-particle interaction effects in the Van Allen belts.

Discovering such structures as the third radiation belt (or "storage ring") has been a major observational achievement of the NASA Radiation Belt Storm Probes program (renamed the "Van Allen Probes" mission in November 2012). A goal of that program was to understand more thoroughly how high-energy electrons are accelerated deep inside the radiation belts-and ultimately lost-due to various wave-particle interactions. Van Allen Probes studies have demonstrated that electrons ranging up to 10 megaelectron volts (MeV) or more can be produced over broad regions of the outer Van Allen zone on timescales as short as a few minutes. The key to such rapid acceleration is the interaction of "seed" populations of ~ 10-200 keV electrons (and subsequently higher energies) with electromagnetic waves in the lower band (whistler-mode) chorus frequency range. Van Allen Probes data show that "source" electrons (in a typical energy range of one to a few tens of keV energy) produced by magnetospheric substorms play a crucial role in feeding free energy into the chorus waves in the outer zone. These chorus waves then, in turn, rapidly heat and accelerate the tens to hundreds of keV seed electrons injected by substorms to much higher energies. Hence, we often see that geomagnetic activity driven by strong solar storms (coronal mass ejections, or CMEs) commonly leads to ultra-relativistic electron production through the intermediary step of waves produced during intense magnetospheric substorms. More generally, wave-particle interactions are of fundamental importance over a broad range of energies and in virtually all regions of the magnetosphere. We provide a summary of many of the wave modes and particle interactions that have been studied in recent times.

Megadepth: efficient coverage quantification for BigWigs and BAMs.

MOTIVATION: A common way to summarize sequencing datasets is to quantify data lying within genes or other genomic intervals. This can be slow and can require different tools for different input file types. RESULTS: Megadepth is a fast tool for quantifying alignments and coverage for BigWig and BAM/CRAM input files, using substantially less memory than the next-fastest competitor. Megadepth can summarize coverage within all disjoint intervals of the Gencode V35 gene annotation for more than 19 000 GTExV8 BigWig files in approximately 1 h using 32 threads. Megadepth is available both as a command-line tool and as an R/Bioconductor package providing much faster quantification compared to the rtracklayer package. AVAILABILITY AND IMPLEMENTATION: https://github.com/ChristopherWilks/megadepth, https://bioconductor.org/packages/megadepth. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

New Insights From Long-Term Measurements of Inner Belt Protons (10s of MeV) by SAMPEX, POES, Van Allen Probes, and Simulation Results.

The Solar, Anomalous, and Magnetospheric Particle Explorer (SAMPEX) mission provided long-term measurements of 10s of megaelectron volt (MeV) inner belt (L < 2) protons (1992-2009) as did the Polar-orbiting Operational Environmental Satellite-18 (POES-18, 2005 to present). These long-term measurements at low-Earth orbit (LEO) showed clear solar cycle variations which anticorrelate with sunspot number. However, the magnitude of the variation is much greater than the solar cycle variation of galactic cosmic rays (>GeV) that are regarded as a source of these trapped protons. Furthermore, the proton fluxes and their variations sensitively depend on the altitude above the South Atlantic Anomaly (SAA) region. With respect to protons (>36 MeV) mirroring near the magnetic equator, both POES measurements and simulations show no obvious solar cycle variations at L > 1.2. This is also confirmed by recent measurements from the Van Allen Probes (2012-2019), but there are clear solar cycle variations and a strong spatial gradient of the proton flux below L = 1.2. A direct comparison between measurements and simulations leads to the conclusion that energy loss of trapped protons due to collisions with free and bound electrons in the ionosphere and atmosphere is the dominant mechanism for the strong spatial gradient and solar cycle variation of the inner belt protons. This fact is also key of importance for spacecraft and instrument design and operation in near-Earth space.

Dashing: fast and accurate genomic distances with HyperLogLog.

Dashing is a fast and accurate software tool for estimating similarities of genomes or sequencing datasets. It uses the HyperLogLog sketch together with cardinality estimation methods that are specialized for set unions and intersections. Dashing summarizes genomes more rapidly than previous MinHash-based methods while providing greater accuracy across a wide range of input sizes and sketch sizes. It can sketch and calculate pairwise distances for over 87K genomes in 6 minutes. Dashing is open source and available at https://github.com/dnbaker/dashing.

Origin of two-band chorus in the radiation belt of Earth.

Naturally occurring chorus emissions are a class of electromagnetic waves found in the space environments of the Earth and other magnetized planets. They play an essential role in accelerating high-energy electrons forming the hazardous radiation belt environment. Chorus typically occurs in two distinct frequency bands separated by a gap. The origin of this two-band structure remains a 50-year old question. Here we report, using NASA's Van Allen Probe measurements, that banded chorus waves are commonly accompanied by two separate anisotropic electron components. Using numerical simulations, we show that the initially excited single-band chorus waves alter the electron distribution immediately via Landau resonance, and suppress the electron anisotropy at medium energies. This naturally divides the electron anisotropy into a low and a high energy components which excite the upper-band and lower-band chorus waves, respectively. This mechanism may also apply to the generation of chorus waves in other magnetized planetary magnetospheres.

Multiyear Measurements of Radiation Belt Electrons: Acceleration, Transport, and Loss.

In addition to clarifying morphological structures of the Earth's radiation belts, it has also been a major achievement of the Van Allen Probes mission to understand more thoroughly how highly relativistic and ultrarelativistic electrons are accelerated deep inside the radiation belts. Prior studies have demonstrated that electrons up to energies of 10 megaelectron volts (MeV) can be produced over broad regions of the outer Van Allen zone on timescales of minutes to a few hours. It often is seen that geomagnetic activity driven by strong solar storms (i.e., coronal mass ejections, or CMEs) almost inexorably leads to relativistic electron production through the intermediary step of intense magnetospheric substorms. In this study, we report observations over the 6-year period 1 September 2012 to 1 September 2018. We focus on data about the relativistic and ultrarelativistic electrons (E≥5 MeV) measured by the Relativistic Electron-Proton Telescope sensors on board the Van Allen Probes spacecraft. This work portrays the radiation belt acceleration, transport, and loss characteristics over a wide range of geomagnetic events. We emphasize features seen repeatedly in the data (three-belt structures, "impenetrable" barrier properties, and radial diffusion signatures) in the context of acceleration and loss mechanisms. We especially highlight solar wind forcing of the ultrarelativistic electron populations and extended periods when such electrons were absent. The analysis includes new display tools showing spatial features of the mission-long time variability of the outer Van Allen belt emphasizing the remarkable dynamics of the system.

Statistical Study of Mercury's Energetic Electron Events as Observed by the Gamma-Ray and Neutron Spectrometer Instrument Onboard MESSENGER.

We present results from a statistical analysis of Mercury's energetic electron (EE) events as observed by the gamma-ray and neutron spectrometer instrument onboard the MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) spacecraft. The main objective of this study is to investigate possible anisotropic behavior of EE events using multiple data sets from MESSENGER instruments. We study the data from the neutron spectrometer (NS) and the gamma-ray spectrometer anticoincidence shield (ACS) because they use the same type of borated plastic scintillator and, hence, they have very similar response functions, and their large surface areas make them more sensitive to low-intensity EE events than MESSENGER's particle instrumentation. The combined analysis of NS and ACS data reveals two different classes of energetic electrons: "Standard" events and "ACS-enhanced" events. Standard events, which comprise over 90% of all events, have signal sizes that are the same in both the ACS and NS. They are likely gyrating particles about Mercury's magnetic field following a 90° pitch angle distribution and are located in well-defined latitude and altitude regions within Mercury's magnetosphere. ACS-enhanced events, which comprise less than 10% of all events, have signal sizes in the ACS that are 10 to 100 times larger than those observed by the NS. They follow a beam-like distribution and are observed both inside and outside Mercury's magnetosphere with a wider range of latitudes and altitudes than Standard events. The difference between the Standard and ACS-enhanced event characteristics suggests distinct underyling acceleration mechanisms.

Measurement of electrons from albedo neutron decay and neutron density in near-Earth space.

The Galaxy is filled with cosmic-ray particles, mostly protons with kinetic energies greater than hundreds of megaelectronvolts. Around Earth, trapped energetic protons, electrons and other particles circulate at altitudes from about 500 to 40,000 kilometres in the Van Allen radiation belts. Soon after these radiation belts were discovered six decades ago, it was recognized that the main source of inner-belt protons (with kinetic energies of tens to hundreds of megaelectronvolts) is cosmic-ray albedo neutron decay (CRAND). In this process, cosmic rays that reach the upper atmosphere interact with neutral atoms to produce albedo neutrons, which, being prone to ß-decay, are a possible source of geomagnetically trapped protons and electrons. These protons would retain most of the kinetic energy of the neutrons, while the electrons would have lower energies, mostly less than one megaelectronvolt. The viability of CRAND as an electron source has, however, been uncertain, because measurements have shown that the electron intensity in the inner Van Allen belt can vary greatly, while the neutron-decay rate should be almost constant. Here we report measurements of relativistic electrons near the inner edge of the inner radiation belt. We demonstrate that the main source of these electrons is indeed CRAND, and that this process also contributes to electrons in the inner belt elsewhere. Furthermore, measurement of the intensity of electrons generated by CRAND provides an experimental determination of the neutron density in near-Earth space-2 × 10-9 per cubic centimetre-confirming theoretical estimates.

Energy-dependent dynamics of keV to MeV electrons in the inner zone, outer zone, and slot regions.

We present observations of the radiation belts from the Helium Oxygen Proton Electron and Magnetic Electron Ion Spectrometer particle detectors on the Van Allen Probes satellites that illustrate the energy dependence and L shell dependence of radiation belt enhancements and decays. We survey events in 2013 and analyze an event on 1 March in more detail. The observations show the following: (a) at all L shells, lower energy electrons are enhanced more often than higher energies; (b) events that fill the slot region are more common at lower energies; (c) enhancements of electrons in the inner zone are more common at lower energies; and (d) even when events do not fully fill the slot region, enhancements at lower energies tend to extend to lower L shells than higher energies. During enhancement events the outer zone extends to lower L shells at lower energies while being confined to higher L shells at higher energies. The inner zone shows the opposite with an outer boundary at higher L shells for lower energies. Both boundaries are nearly straight in log(energy) versus L shell space. At energies below a few 100 keV, radiation belt electron penetration through the slot region into the inner zone is commonplace, but the number and frequency of "slot filling" events decreases with increasing energy. The inner zone is enhanced only at energies that penetrate through the slot. Energy- and L shell-dependent losses (that are consistent with whistler hiss interactions) return the belts to more quiescent conditions.

Intense energetic electron flux enhancements in Mercury's magnetosphere: An integrated view with high-resolution observations from MESSENGER.

The MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) mission to Mercury has provided a wealth of new data about energetic particle phenomena. With observations from MESSENGER's Energetic Particle Spectrometer, as well as data arising from energetic electrons recorded by the X-Ray Spectrometer and Gamma-Ray and Neutron Spectrometer (GRNS) instruments, recent work greatly extends our record of the acceleration, transport, and loss of energetic electrons at Mercury. The combined data sets include measurements from a few keV up to several hundred keV in electron kinetic energy and have permitted relatively good spatial and temporal resolution for many events. We focus here on the detailed nature of energetic electron bursts measured by the GRNS system, and we place these events in the context of solar wind and magnetospheric forcing at Mercury. Our examination of data at high temporal resolution (10 ms) during the period March 2013 through October 2014 supports strongly the view that energetic electrons are accelerated in the near-tail region of Mercury's magnetosphere and are subsequently "injected" onto closed magnetic field lines on the planetary nightside. The electrons populate the plasma sheet and drift rapidly eastward toward the dawn and prenoon sectors, at times executing multiple complete drifts around the planet to form "quasi-trapped" populations.

Wave-induced loss of ultra-relativistic electrons in the Van Allen radiation belts.

The dipole configuration of the Earth's magnetic field allows for the trapping of highly energetic particles, which form the radiation belts. Although significant advances have been made in understanding the acceleration mechanisms in the radiation belts, the loss processes remain poorly understood. Unique observations on 17 January 2013 provide detailed information throughout the belts on the energy spectrum and pitch angle (angle between the velocity of a particle and the magnetic field) distribution of electrons up to ultra-relativistic energies. Here we show that although relativistic electrons are enhanced, ultra-relativistic electrons become depleted and distributions of particles show very clear telltale signatures of electromagnetic ion cyclotron wave-induced loss. Comparisons between observations and modelling of the evolution of the electron flux and pitch angle show that electromagnetic ion cyclotron waves provide the dominant loss mechanism at ultra-relativistic energies and produce a profound dropout of the ultra-relativistic radiation belt fluxes.

Fragment Length of Circulating Tumor DNA.

Malignant tumors shed DNA into the circulation. The transient half-life of circulating tumor DNA (ctDNA) may afford the opportunity to diagnose, monitor recurrence, and evaluate response to therapy solely through a non-invasive blood draw. However, detecting ctDNA against the normally occurring background of cell-free DNA derived from healthy cells has proven challenging, particularly in non-metastatic solid tumors. In this study, distinct differences in fragment length size between ctDNAs and normal cell-free DNA are defined. Human ctDNA in rat plasma derived from human glioblastoma multiforme stem-like cells in the rat brain and human hepatocellular carcinoma in the rat flank were found to have a shorter principal fragment length than the background rat cell-free DNA (134-144 bp vs. 167 bp, respectively). Subsequently, a similar shift in the fragment length of ctDNA in humans with melanoma and lung cancer was identified compared to healthy controls. Comparison of fragment lengths from cell-free DNA between a melanoma patient and healthy controls found that the BRAF V600E mutant allele occurred more commonly at a shorter fragment length than the fragment length of the wild-type allele (132-145 bp vs. 165 bp, respectively). Moreover, size-selecting for shorter cell-free DNA fragment lengths substantially increased the EGFR T790M mutant allele frequency in human lung cancer. These findings provide compelling evidence that experimental or bioinformatic isolation of a specific subset of fragment lengths from cell-free DNA may improve detection of ctDNA.

Near-Earth injection of MeV electrons associated with intense dipolarization electric fields: Van Allen Probes observations.

Substorms generally inject tens to hundreds of keV electrons, but intense substorm electric fields have been shown to inject MeV electrons as well. An intriguing question is whether such MeVelectron injections can populate the outer radiation belt. Here we present observations of a substorm injection of MeV electrons into the inner magnetosphere. In the premidnight sector at L ∼ 5.5, Van Allen Probes (Radiation Belt Storm Probes)-A observed a large dipolarization electric field (50 mV/m) over ∼40 s and a dispersionless injection of electrons up to ∼3 MeV. Pitch angle observations indicated betatron acceleration of MeV electrons at the dipolarization front. Corresponding signals of MeV electron injection were observed at LANL-GEO, THEMIS-D, and GOES at geosynchronous altitude. Through a series of dipolarizations, the injections increased the MeV electron phase space density by 1 order of magnitude in less than 3 h in the outer radiation belt (L > 4.8). Our observations provide evidence that deep injections can supply significant MeV electrons.

MESSENGER observations of the spatial distribution of planetary ions near Mercury.

Global measurements by MESSENGER of the fluxes of heavy ions at Mercury, particularly sodium (Na(+)) and oxygen (O(+)), exhibit distinct maxima in the northern magnetic-cusp region, indicating that polar regions are important sources of Mercury's ionized exosphere, presumably through solar-wind sputtering near the poles. The observed fluxes of helium (He(+)) are more evenly distributed, indicating a more uniform source such as that expected from evaporation from a helium-saturated surface. In some regions near Mercury, especially the nightside equatorial region, the Na(+) pressure can be a substantial fraction of the proton pressure.