Most cancer genomic data are generated from bulk samples composed of mixtures of cancer subpopulations, as well as normal cells. Subclonal reconstruction methods based on machine learning aim to separate those subpopulations in a sample and infer their evolutionary history. However, current approaches are entirely data driven and agnostic to evolutionary theory. We demonstrate that systematic errors occur in the analysis if evolution is not accounted for, and this is exacerbated with multi-sampling of the same tumor. We present a novel approach for model-based tumor subclonal reconstruction, called MOBSTER, which combines machine learning with theoretical population genetics. Using public whole-genome sequencing data from 2,606 samples from different cohorts, new data and synthetic validation, we show that this method is more robust and accurate than current techniques in single-sample, multiregion and longitudinal data. This approach minimizes the confounding factors of nonevolutionary methods, thus leading to more accurate recovery of the evolutionary history of human cancers.Although hormones such as glucocorticoids have been broadly accepted in recent decades as general neuromodulators of memory processes, sex steroid hormones such as the potent oestrogen 17β-oestradiol have been less well recognized by the scientific community in this capacity. The predominance of females in studies of oestradiol and memory and the general (but erroneous) perception that oestrogens are ‘female’ hormones have probably prevented oestradiol from being more widely considered as a key memory modulator in both sexes. Indeed, although considerable evidence supports a crucial role for oestradiol in regulating learning and memory in females, a growing body of literature indicates a similar role in males. This Review discusses the mechanisms of oestradiol signalling and provides an overview of the effects of oestradiol on spatial, object recognition, social and fear memories. Although the primary focus is on data collected in females, effects of oestradiol on memory in males will be discussed, as will sex differences in the molecular mechanisms that regulate oestrogenic modulation of memory, which may have important implications for the development of future cognitive therapeutics.The first issue of Nature Reviews Neuroscience was published 20 years ago, in 2000. To mark this anniversary, in this Viewpoint article we asked a selection of researchers from across the field who have authored pieces published in the journal in recent years for their thoughts on notable and interesting developments in neuroscience, and particularly in their areas of the field, over the past two decades. They also provide some thoughts on current lines of research and questions that excite them.The masses of the lightest atomic nuclei and the electron mass1 are interlinked, and their values affect observables in atomic2, molecular3-5 and neutrino physics6, as well as metrology. The most precise values for these fundamental parameters come from Penning trap mass spectrometry, which achieves relative mass uncertainties of the order of 10-11. However, redundancy checks using data from different experiments reveal considerable inconsistencies in the masses of the proton, the deuteron and the helion (the nucleus of helium-3), suggesting that the uncertainty of these values may have been underestimated. Here we present results from absolute mass measurements of the deuteron and the HD+ molecular ion using 12C as a mass reference. Our value for the deuteron mass, 2.013553212535(17) atomic mass units, has better precision than the CODATA value7 by a factor of 2.4 and differs from it by 4.8 standard deviations. With a relative uncertainty of eight parts per trillion, this is the most precise mass value measured directly in atomic mass units. Furthermore, our measurement of the mass of the HD+ molecular ion, 3.021378241561(61) atomic mass units, not only allows a rigorous consistency check of our results for the masses of the deuteron (this work) and the proton8, but also establishes an additional link for the masses of tritium9 and helium-3 (ref. 10) to the atomic mass unit. Combined with a recent measurement of the deuteron-to-proton mass ratio11, the uncertainty of the reference value of the proton mass7 can be reduced by a factor of three.When placed over a less dense medium, a liquid layer will typically collapse downwards if it exceeds a certain size, as gravity acting on the lower liquid interface triggers a destabilizing effect called a Rayleigh-Taylor instability1,2. Of the many methods that have been developed to prevent the liquid from falling3-6, vertical shaking has proved to be efficient and has therefore been studied in detail7-13. Stabilization is the result of the dynamical averaging effect of the oscillating effective gravity. Vibrations of liquids also induce other paradoxical phenomena such as the sinking of air bubbles14-19 or the stabilization of heavy objects in columns of fluid at unexpected heights20. Here we take advantage of the excitation resonance of the supporting air layer to perform experiments with large levitating liquid layers of up to half a litre in volume and up to 20 centimetres in width. Moreover, we predict theoretically and show experimentally that vertical shaking also creates stable buoyancy positions on the lower interface of the liquid, which behave as though the gravitational force were inverted. Bodies can thus float upside down on the lower interface of levitating liquid layers. We use our model to predict the minimum excitation needed to withstand falling of such an inverted floater, which depends on its mass. Experimental observations confirm the possibility of selective falling of heavy bodies. Our findings invite us to rethink all interfacial phenomena in this exotic and counter-intuitive stable configuration.Rechargeable lithium-ion batteries with high energy density that can be safely charged and discharged at high rates are desirable for electrified transportation and other applications1-3. However, the sub-optimal intercalation potentials of current anodes result in a trade-off between energy density, power and safety. Here we report that disordered rock salt4,5 Li3+xV2O5 can be used as a fast-charging anode that can reversibly cycle two lithium ions at an average voltage of about 0.6 volts versus a Li/Li+ reference electrode. The increased potential compared to graphite6,7 reduces the likelihood of lithium metal plating if proper charging controls are used, alleviating a major safety concern (short-circuiting related to Li dendrite growth). GSK2795039 In addition, a lithium-ion battery with a disordered rock salt Li3V2O5 anode yields a cell voltage much higher than does a battery using a commercial fast-charging lithium titanate anode or other intercalation anode candidates (Li3VO4 and LiV0.5Ti0.5S2)8,9. Further, disordered rock salt Li3V2O5 can perform over 1,000 charge-discharge cycles with negligible capacity decay and exhibits exceptional rate capability, delivering over 40 per cent of its capacity in 20 seconds.