BITES: balanced individual treatment effect for survival data.

MOTIVATION: Estimating the effects of interventions on patient outcome is one of the key aspects of personalized medicine. Their inference is often challenged by the fact that the training data comprises only the outcome for the administered treatment, and not for alternative treatments (the so-called counterfactual outcomes). Several methods were suggested for this scenario based on observational data, i.e. data where the intervention was not applied randomly, for both continuous and binary outcome variables. However, patient outcome is often recorded in terms of time-to-event data, comprising right-censored event times if an event does not occur within the observation period. Albeit their enormous importance, time-to-event data are rarely used for treatment optimization. We suggest an approach named BITES (Balanced Individual Treatment Effect for Survival data), which combines a treatment-specific semi-parametric Cox loss with a treatment-balanced deep neural network; i.e. we regularize differences between treated and non-treated patients using Integral Probability Metrics (IPM). RESULTS: We show in simulation studies that this approach outperforms the state of the art. Furthermore, we demonstrate in an application to a cohort of breast cancer patients that hormone treatment can be optimized based on six routine parameters. We successfully validated this finding in an independent cohort. AVAILABILITY AND IMPLEMENTATION: We provide BITES as an easy-to-use python implementation including scheduled hyper-parameter optimization (https://github.com/sschrod/BITES). The data underlying this article are available in the CRAN repository at https://rdrr.io/cran/survival/man/gbsg.html and https://rdrr.io/cran/survival/man/rotterdam.html. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Pyrobaculum yellowstonensis Strain WP30 Respires on Elemental Sulfur and/or Arsenate in Circumneutral Sulfidic Geothermal Sediments of Yellowstone National Park.

Thermoproteales (phylum Crenarchaeota) populations are abundant in high-temperature (>70°C) environments of Yellowstone National Park (YNP) and are important in mediating the biogeochemical cycles of sulfur, arsenic, and carbon. The objectives of this study were to determine the specific physiological attributes of the isolate Pyrobaculum yellowstonensis strain WP30, which was obtained from an elemental sulfur sediment (Joseph's Coat Hot Spring [JCHS], 80°C, pH 6.1, 135 µM As) and relate this organism to geochemical processes occurring in situ. Strain WP30 is a chemoorganoheterotroph and requires elemental sulfur and/or arsenate as an electron acceptor. Growth in the presence of elemental sulfur and arsenate resulted in the formation of thioarsenates and polysulfides. The complete genome of this organism was sequenced (1.99 Mb, 58% G+C content), revealing numerous metabolic pathways for the degradation of carbohydrates, amino acids, and lipids. Multiple dimethyl sulfoxide-molybdopterin (DMSO-MPT) oxidoreductase genes, which are implicated in the reduction of sulfur and arsenic, were identified. Pathways for the de novo synthesis of nearly all required cofactors and metabolites were identified. The comparative genomics of P. yellowstonensis and the assembled metagenome sequence from JCHS showed that this organism is highly related (∼95% average nucleotide sequence identity) to in situ populations. The physiological attributes and metabolic capabilities of P. yellowstonensis provide an important foundation for developing an understanding of the distribution and function of these populations in YNP.

Anaerobic Chemolithotrophic Growth of the Haloalkaliphilic Bacterium Strain MLMS-1 by Disproportionation of Monothioarsenate.

A novel chemolithotrophic metabolism based on a mixed arsenic-sulfur species has been discovered for the anaerobic deltaproteobacterium, strain MLMS-1, a haloalkaliphile isolated from Mono Lake, California, U.S. Strain MLMS-1 is the first reported obligate arsenate-respiring chemoautotroph which grows by coupling arsenate reduction to arsenite with the oxidation of sulfide to sulfate. In that pathway the formation of a mixed arsenic-sulfur species was reported. That species was assumed to be monothioarsenite ([H2As(III)S(-II)O2](-)), formed as an intermediate by abiotic reaction of arsenite with sulfide. We now report that this species is monothioarsenate ([HAs(V)S(-II)O3](2-)) as revealed by X-ray absorption spectroscopy. Monothioarsenate forms by abiotic reaction of arsenite with zerovalent sulfur. Monothioarsenate is kinetically stable under a wide range of pH and redox conditions. However, it was metabolized rapidly by strain MLMS-1 when incubated with arsenate. Incubations using monothioarsenate confirmed that strain MLMS-1 was able to grow (µ = 0.017 h(-1)) on this substrate via a disproportionation reaction by oxidizing the thio-group-sulfur (S(-II)) to zerovalent sulfur or sulfate while concurrently reducing the central arsenic atom (As(V)) to arsenite. Monothioarsenate disproportionation could be widespread in nature beyond the already studied arsenic and sulfide rich hot springs and soda lakes where it was discovered.

Nonanalyticity in scale in the planar limit of QCD.

Using methods of numerical lattice gauge theory we show that, in the limit of a large number of colors, properly regularized Wilson loops have an eigenvalue distribution which changes nonanalytically as the overall size of the loop is increased. This establishes a large-N phase transition in continuum planar gauge theory, a fact whose precise implications remain to be worked out.