It has been shown that news events influence the trends of stock price movements. However, previous work on news-driven stock market prediction rely on shallow features (such as bags-of-words, named entities and noun phrases), which do not capture structured entity-relation information, and hence cannot represent complete and exact events. Recent advances in Open Information Extraction (Open IE) techniques enable the extraction of structured events from web-scale data. We propose to adapt Open IE technology for event-based stock price movement prediction, extracting structured events from large-scale public news without manual efforts. Both linear and nonlinear models are employed to empirically investigate the hidden and complex relationships between events and the stock market. Largescale experiments show that the accuracy of S&P 500 index prediction is 60%, and that of individual stock prediction can be over 70%. Our event-based system outperforms bags-of-words-based baselines, and previously reported systems trained on S&P 500 stock historical data.

No abstract

DM1, a maytansine derivative, is a highly potential cytotoxic agent but with severe side effects; therefore, its application in clinical cancer therapy is limited. Here, in order to mitigate this intrinsic drawback of DM1, we developed mesoporous silica nanoparticles (MSNs) loaded with DM1 and surface-decorated with hydrochloride dopamine (PDA), polyethylene glycol (PEG), and epithelial cell adhesion molecule (EpCAM) aptamer (APt) for the targeted treatment of colorectal cancer (CRC). In this system, the PDA coating could be used as pH-sensitive gatekeepers to control the release of DM1 from MSNs in response to the pH stimulus and EpCAM APt-guided active targeting enables the increased delivery of DM1 to CRC as well as a reduction in toxicity and side effects by minimizing the exposure of normal tissues to DM1. Results demonstrated that DM1 inhibited the formation of microtubules and induced apoptosis in tumor cells via caspase signaling. In comparison with the control groups, the MSNs-DM1@PDA-PEG-APt bioconjugates exhibited increased binding ability and much higher cytotoxicity to the CRC SW480 cell line. Furthermore, in vivo assays confirmed the advantages of such a strategy. These findings suggested that MSNs-DM1@PDA-PEG-APt could represent a promising therapeutic platform for EpCAM-positive CRC.

It was and remains a big challenge for cancer nanomedicines to achieve high and stable drug loading with fast drug release in the target cells. Here, we report on novel hyaluronic acid-shelled disulfide-cross-linked biodegradable polymersomes (HA-XPS) self-assembled from hyaluronic acid-b-poly(trimethylene carbonate-co-dithiolane trimethylene carbonate) diblock copolymer for ultrahigh-efficiency reactive encapsulation and CD44-targeted delivery of mertansine (DM1) toxin, a highly potent warhead for clinically used antibody-drug conjugates. Remarkably, HA-XPS showed quantitative encapsulation of DM1 even with a high drug loading content of 16.7 wt %. DM1-loaded HA-XPS (HA-XPS-DM1) presented a small size of ∼80 nm, low drug leakage under physiological conditions, and fast glutathione-triggered drug release. MTT assays revealed that HA-XPS was noncytotoxic while HA-XPS-DM1 was highly potent to MDA-MB-231 cells with an IC comparable to that of free DM1. The in vitro and in vivo inhibition experiments indicated that HA-XPS could actively target MDA-MB-231 cells. Notably, HA-XPS-DM1 while causing little adverse effect could effectively inhibit tumor growth and significantly prolong survival time in MDA-MB-231 human breast tumor-bearing mice. HA-XPS-DM1 provides a novel and unique treatment for CD44-positive cancers.

No abstract