Machine learning advances chemistry and materials science by enabling large-scale exploration of chemical space based on quantum chemical calculations. While these models supply fast and accurate ...predictions of atomistic chemical properties, they do not explicitly capture the electronic degrees of freedom of a molecule, which limits their applicability for reactive chemistry and chemical analysis. Here we present a deep learning framework for the prediction of the quantum mechanical wavefunction in a local basis of atomic orbitals from which all other ground-state properties can be derived. This approach retains full access to the electronic structure via the wavefunction at force-field-like efficiency and captures quantum mechanics in an analytically differentiable representation. On several examples, we demonstrate that this opens promising avenues to perform inverse design of molecular structures for targeting electronic property optimisation and a clear path towards increased synergy of machine learning and quantum chemistry.
Nonadiabatic effects in chemical reaction at metal surfaces, due to excitation of electron–hole pairs, stand at the frontier of the studies of gas-surface reaction dynamics. However, the ...first-principles description of electronic excitation remains challenging. In an efficient molecular dynamics with electronic friction (MDEF) method, the nonadiabatic couplings are effectively included in a so-called electronic friction tensor (EFT), which can be computed from first-order time-dependent perturbation theory (TDPT) in terms of density functional theory (DFT) orbitals. This second-rank tensor depends on adsorbate position and features a complicated transformation with regard to the intrinsic symmetry operations of the system. In this work, we develop a new symmetry-adapted neural network representation of EFT, based on our recently proposed embedded atom neural network (EANN) framework. Inspired by the derivation of the nonadiabatic coupling matrix, we represent the tensorial friction by the first and second derivatives of multiple outputs of NNs with respect to atomic Cartesian coordinates. This rigorously preserves the positive semidefiniteness, directional property, and correct symmetry-equivariance of EFT. Unlike previous methods, our new approach can readily include both molecular and surface degrees of freedom, regardless of the type of surface. Tests on the H2 + Ag(111) system show that this approach yields an accurate, efficient, and continuous representation of EFT, making it possible to perform large scale TDPT-based MDEF simulations to study both adiabatic and nonadiabatic energy dissipation in a unified framework.
Himalayan glaciers supply meltwater to densely populated catchments in South Asia, and regional observations of glacier change over multiple decades are needed to understand climate drivers and ...assess resulting impacts on glacier-fed rivers. Here, we quantify changes in ice thickness during the intervals 1975-2000 and 2000-2016 across the Himalayas, using a set of digital elevation models derived from cold war-era spy satellite film and modern stereo satellite imagery. We observe consistent ice loss along the entire 2000-km transect for both intervals and find a doubling of the average loss rate during 2000-2016 -0.43 ± 0.14 m w.e. year
(meters of water equivalent per year) compared to 1975-2000 (-0.22 ± 0.13 m w.e. year
). The similar magnitude and acceleration of ice loss across the Himalayas suggests a regionally coherent climate forcing, consistent with atmospheric warming and associated energy fluxes as the dominant drivers of glacier change.
Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma. Although 5-year survival rates in the first-line setting range from 60% to 70%, up to 50% of patients become ...refractory to or relapse after treatment. Published analyses of large-scale outcome data from patients with refractory DLBCL are limited. SCHOLAR-1, an international, multicohort retrospective non-Hodgkin lymphoma research study, retrospectively evaluated outcomes in patients with refractory DLBCL which, for this study, was defined as progressive disease or stable disease as best response at any point during chemotherapy (>4 cycles of first-line or 2 cycles of later-line therapy) or relapsed at ≤12 months from autologous stem cell transplantation. SCHOLAR-1 pooled data from 2 phase 3 clinical trials (Lymphoma Academic Research Organization-CORAL and Canadian Cancer Trials Group LY.12) and 2 observational cohorts (MD Anderson Cancer Center and University of Iowa/Mayo Clinic Lymphoma Specialized Program of Research Excellence). Response rates and overall survival were estimated from the time of initiation of salvage therapy for refractory disease. Among 861 patients, 636 were included on the basis of refractory disease inclusion criteria. For patients with refractory DLBCL, the objective response rate was 26% (complete response rate, 7%) to the next line of therapy, and the median overall survival was 6.3 months. Twenty percent of patients were alive at 2 years. Outcomes were consistently poor across patient subgroups and study cohorts. SCHOLAR-1 is the largest patient-level pooled retrospective analysis to characterize response rates and survival for a population of patients with refractory DLBCL.
•SCHOLAR-1 is the first patient-level analysis of outcomes of refractory DLBCL from 2 large randomized trials and 2 academic databases.•SCHOLAR-1 demonstrated poor outcomes in patients with refractory DLBCL, supporting a need for more effective therapies for these patients.
During the past two decades, density-functional (DF) theory has evolved from niche applications for simple solid-state materials to become a workhorse method for studying a wide range of phenomena in ...a variety of system classes throughout physics, chemistry, biology, and materials science. Here, we review the recent advances in DF calculations for materials modeling, giving a classification of modern DF-based methods when viewed from the materials modeling perspective. While progress has been very substantial, many challenges remain on the way to achieving consensus on a set of universally applicable DF-based methods for materials modeling. Hence, we focus on recent successes and remaining challenges in DF calculations for modeling hard solids, molecular and biological matter, low-dimensional materials, and hybrid organic-inorganic materials.
Twenty percent of patients with follicular lymphoma (FL) experience progression of disease (POD) within 2 years of initial chemoimmunotherapy. We analyzed data from the National LymphoCare Study to ...identify whether prognostic FL factors are associated with early POD and whether patients with early POD are at high risk for death.
In total, 588 patients with stage 2 to 4 FL received first-line rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP). Two groups were defined: patients with early POD 2 years or less after diagnosis and those without POD within 2 years, the reference group. An independent validation set, 147 patients with FL who received first-line R-CHOP, was analyzed for reproducibility.
Of 588 patients, 19% (n = 110) had early POD, 71% (n = 420) were in the reference group, 8% (n = 46) were lost to follow-up, and 2% (n = 12) died without POD less than 2 years after diagnosis. Five-year overall survival was lower in the early-POD group than in the reference group (50% v 90%). This trend was maintained after we adjusted for FL International Prognostic Index (hazard ratio, 6.44; 95% CI, 4.33 to 9.58). Results were similar for the validation set (FL International Prognostic Index-adjusted hazard ratio, 19.8).
In patients with FL who received first-line R-CHOP, POD within 2 years after diagnosis was associated with poor outcomes and should be further validated as a standard end point of chemoimmunotherapy trials of untreated FL. This high-risk FL population warrants further study in directed prospective clinical trials.