A Causal Framework for Distribution Generalization.

We consider the problem of predicting a response Y from a set of covariates X when test- and training distributions differ. Since such differences may have causal explanations, we consider test distributions that emerge from interventions in a structural causal model, and focus on minimizing the worst-case risk. Causal regression models, which regress the response on its direct causes, remain unchanged under arbitrary interventions on the covariates, but they are not always optimal in the above sense. For example, for linear models and bounded interventions, alternative solutions have been shown to be minimax prediction optimal. We introduce the formal framework of distribution generalization that allows us to analyze the above problem in partially observed nonlinear models for both direct interventions on X and interventions that occur indirectly via exogenous variables A. It takes into account that, in practice, minimax solutions need to be identified from data. Our framework allows us to characterize under which class of interventions the causal function is minimax optimal. We prove sufficient conditions for distribution generalization and present corresponding impossibility results. We propose a practical method, NILE, that achieves distribution generalization in a nonlinear IV setting with linear extrapolation. We prove consistency and present empirical results.

The three major axes of terrestrial ecosystem function.

The leaf economics spectrum1,2 and the global spectrum of plant forms and functions3 revealed fundamental axes of variation in plant traits, which represent different ecological strategies that are shaped by the evolutionary development of plant species2. Ecosystem functions depend on environmental conditions and the traits of species that comprise the ecological communities4. However, the axes of variation of ecosystem functions are largely unknown, which limits our understanding of how ecosystems respond as a whole to anthropogenic drivers, climate and environmental variability4,5. Here we derive a set of ecosystem functions6 from a dataset of surface gas exchange measurements across major terrestrial biomes. We find that most of the variability within ecosystem functions (71.8%) is captured by three key axes. The first axis reflects maximum ecosystem productivity and is mostly explained by vegetation structure. The second axis reflects ecosystem water-use strategies and is jointly explained by variation in vegetation height and climate. The third axis, which represents ecosystem carbon-use efficiency, features a gradient related to aridity, and is explained primarily by variation in vegetation structure. We show that two state-of-the-art land surface models reproduce the first and most important axis of ecosystem functions. However, the models tend to simulate more strongly correlated functions than those observed, which limits their ability to accurately predict the full range of responses to environmental changes in carbon, water and energy cycling in terrestrial ecosystems7,8.

Cytokine Profile in Patients with Aseptic Loosening of Total Hip Replacements and Its Relation to Metal Release and Metal Allergy.

Metal release from total hip replacements (THRs) is associated with aseptic loosening (AL). It has been proposed that the underlying immunological response is caused by a delayed type IV hypersensitivity-like reaction to metals, i.e., metal allergy. The purpose of this study was to investigate the immunological response in patients with AL in relation to metal release and the prevalence of metal allergy. THR patients undergoing revision surgery due to AL or mechanical implant failures were included in the study along with a control group consisting of primary THR patients. Comprehensive cytokine analyses were performed on serum and periimplant tissue samples along with metal analysis using inductive coupled plasma mass spectrometry (ICP-MS). Patient patch testing was done with a series of metals related to orthopedic implant. A distinct cytokine profile was found in the periimplant tissue of patients with AL. Significantly increased levels of the proinflammatory cytokines IL-1ß, IL-2, IL-8, IFN-γ and TNF-α, but also the anti-inflammatory IL-10 were detected. A general increase of metal concentrations in the periimplant tissue was observed in both revision groups, while Cr was significantly increased in patient serum with AL. No difference in the prevalence of metal sensitivity was established by patch testing. Increased levels of IL-1ß, IL-8, and TNF-α point to an innate immune response. However, the presence of IL-2 and IFN-γ indicates additional involvement of T cell-mediated response in patients with AL, although this could not be detected by patch testing.

Subcoronary Stentless Aortic Valves are Not Superior to Supra-Annular Stented Valves Regarding Turbulent Stress.

BACKGROUND: Regions of turbulence downstream of bioprosthetic heart valves may cause damage to blood components, vessel walls, and also to aortic valve leaflets. Stentless aortic heart valves are known to possess several hemodynamic benefits such as a larger effective orifice area and a lower aortic transvalvular pressure difference compared to their stented counterparts. To date, turbulence analysis downstream of a stentless valve prosthesis has been investigated exclusively indirectly, using magnetic resonance imaging or in animal settings only. The study aim was to investigate turbulence using direct Doppler ultrasonography measurements in subcoronary stentless and stented valves in human subjects. METHODS: Either stented pericardial valve prostheses (Mitroflow) or stentless valve prostheses (Solo) were implanted in 15 patients in a randomized fashion. Following surgery, blood velocity was measured in the cross-sectional area downstream of the valves using 10 MHz ultrasonic probes connected to dedicated pulsed Doppler equipment. As a measure of turbulence, Reynolds normal stress (RNS) values were calculated, as well as two-dimensional maps of the turbulence distribution. Preoperative and perioperative data were collected prospectively, and postoperative data retrospectively, from hospital records. RESULTS: The median follow up was 1,624 days. No differences were found in perioperative or postoperative clinical data. Implantation of the Mitroflow valve was significantly faster than that of the Solo valve (p <0.05). Neither was any difference found in the mean or max RNS between the two valve groups. However, the turbulence profiles showed a large variation in the Solo valve compared to the more uniform profiles of the Mitroflow valve. CONCLUSIONS: Comparable turbulent flow values were found between the two valve types, although the Solo group exhibited a large variation in turbulence profiles. As no clear clinical advantages were shown to exist for stentless valves, a normal stented valve should be the first choice in most cases.

Differential regulation of the OASL and OAS1 genes in response to viral infections.

The 2'-5' oligoadenylate synthetase (OAS) family consist of three genes encoding active OAS enzymes (OAS1-3) and an OAS-Like (OASL) gene encoding an inactive protein. The transcription of all four members of this family is actively induced by interferon (IFN), but so far no attempt to systematically analyze the expression of these genes during viral infection has been made. We analyzed the expression of the human OAS1 and OASL genes in response to infection with Sendai virus or Influenza A virus. Surprisingly, we found a marked difference in the expression pattern of these genes. Our data showed that the OASL gene is rapidly induced in response to viral infection and that this induction is mediated by IFN regulatory factor 3 (IRF-3). In contrast to the OASL gene, the induction of the OAS1 gene by virus infection was lower, and did require a functional type I IFN response. The pronounced difference in gene regulation between the OAS1 and OASL genes agrees with a functional difference between these genes, which must exist as a consequence of the lack of the 2-5A synthetase activity of the OASL protein. Furthermore, the behavior of the OASL gene is consistent with the behavior of an antiviral gene.