Search

BackgroundRNA features a highly negatively charged phosphate backbone that attracts a of cloud counter-ions that reduce the electrostatic repulsion in a concentration dependent manner. Ion concentrations thus have a large influence on folding and stability of RNA structures. Despite their well-documented effects, salt effects are not handled by currently available secondary stucture prediction algorithms. Combining Debye-Hückel potentials for line charges and Manning’s counter-ion condensation theory, Einertet al. [Biophys. J.100: 2745-2753 (2011)] modeled the energetic effects contributions monovalent cations on loops and helices.ResultsThe model of Einertet al. is adapted to match the structure of the dynamic programming recursion of RNA secondary structure prediction algorithms. An empirical term describing the dependence salt dependence of the duplex initiation energy is added to improve co-folding predictions for two or more RNA strands. The slightly modified model is implemented in theViennaRNApackage in such way that only the energy parameters but not the algorithmic structure is affected. A comparison with data from the literature show that predicted free energies and melting temperatures are in reasonable agreement with experiments.ConclusionThe new feature in theViennaRNApackage makes it possible to study effects of salt concentrations on RNA folding in a systematic manner. Strictly speaking, the model pertains only to mono-valent cations, and thus covers the most important parameter, i.e., the NaCl concentration. It remains a question for future research to what extent unspecific effects of bi- and tri-valent cations can be approximated in a similar manner.AvailabilityCorrections for the concentration of monovalent cations are available in theViennaRNApackage starting from version 2.6.0.

Objectives To compare the clinical outcomes between intravenous carboplatin/paclitaxel chemotherapy plus bevacizumab versus intraperitoneal cisplatin/paclitaxel chemotherapy without bevacizumab as the frontline treatment in women with advanced ovarian, fallopian tube and primary peritoneal cancer. Methods All consecutive women with stage II~IV cancer treated with either frontline intraperitoneal cisplatin/paclitaxel without bevacizumab (IP group) or intravenous carboplatin/paclitaxel with bevacizumab (IVB group) at a tertiary referral center were reviewed. Results A total of 59 women (IP group, n=44; IVB group, n=15) were reviewed. There was no significant difference in the progression-free survival (median: 33.6 versus 14.8 months, p=0.13). However, overall survival (OS) was significantly higher in the IP group, compared with the IVB group (median: not reached versus 31.7 months, p=0.02; adjusted hazard ratio (HR)=0.35, 95% confidence interval (CI)=0.10 to 1.07, p=0.065). Additional predictors for OS include cancer stage and the number of chemotherapy cycles. Besides, the standard dose of 100 mg/m2 cisplatin was a predictor for OS, compared with other intraperitoneal regimens (adjusted HR=0.14, 95% CI=0.02 to 0.87, p=0.03). Conclusions Intraperitoneal cisplatin/paclitaxel chemotherapy without bevacizumab seems to be better in OS, compared to intravenous carboplatin/paclitaxel chemotherapy with bevacizumab in the frontline treatment of women with advanced ovarian cancer.

Prompt learning has been proven to be highly effective in improving pre-trained language model (PLM) adaptability, surpassing conventional fine-tuning paradigms, and showing exceptional promise in an ever-growing landscape of applications and APIs tailored for few-shot learning scenarios. Despite the growing prominence of prompt learning-based APIs, their security concerns remain underexplored. In this paper, we undertake a pioneering study on the Trojan susceptibility of prompt-learning PLM APIs. We identified several key challenges, including discrete-prompt, few-shot, and black-box settings, which limit the applicability of existing backdoor attacks. To address these challenges, we propose TrojPrompt, an automatic and black-box framework to effectively generate universal and stealthy triggers and insert Trojans into hard prompts. Specifically, we propose a universal API-driven trigger discovery algorithm for generating universal triggers for various inputs by querying victim PLM APIs using few-shot data samples. Furthermore, we introduce a novel progressive trojan poisoning algorithm designed to generate poisoned prompts that retain efficacy and transferability across a diverse range of models. Our experiments and results demonstrate TrojPrompt's capacity to effectively insert Trojans into text prompts in real-world black-box PLM APIs, while maintaining exceptional performance on clean test sets and significantly outperforming baseline models. Our work sheds light on the potential security risks in current models and offers a potential defensive approach., Comment: 14 pages, 4 figures, 5 tables

Mathematical models for stream depletion with stream stage decline or drawdown are developed to overcome the deficiency in existing models that typically use the constant-head (Dirichlet) or general (Robin) boundary condition and source terms at the stream-aquifer interface. Existing approaches assume a fixed stream stage during pumping, implies that the stream is an infinite water source, with depletion defined as a decrease in stream discharge. We refer to this depletion without drawdown as the ``stream depletion paradox.'' It is a glaring model limitation, ignoring the most observable adverse effect of long-term groundwater abstraction near a stream, namely stage declines that eventually lead to dry streambeds. Field data are presented to demonstrate that stream stage responds to pumping near the stream, motivating the development of an alternative theory predicts transient stream drawdown based on the concepts of finite stream storage and mass continuity at the stream-aquifer interface. Based on this alternative theory, models are developed for the cases of a non- and a fully-penetrating stream. The proposed model reduces to the fixed-stage model in the limit as stream storage becomes infinitely large and to the limiting case of confined aquifer flow with a no-flow boundary at the streambed when the stream storage vanishes. The model is applied to field observations of both aquifer and stream drawdown from tests conducted in a confined aquifer over which a shallow stream flows. Model fits and parameter estimates are obtained both aquifer and stream drawdown data. Model predicted and observed transient drawdown behavior indicate that fixed-stage models (a) underestimate late-time aquifer drawdown and (b) overestimate the available recharge from streams to pumping wells. This has significant implications for the sustainable management of water resources in hydraulically connected stream-aquifer systems with heavy groundwater abstraction.