Dataless text classification has attracted increasing attentions recently. It only needs very few seed words of each category to classify documents, which is much cheaper than supervised text classification that requires massive labeling efforts. However, most of existing models pay attention to long texts, but get unsatisfactory performance on short texts, which have become increasingly popular on the Internet. In this paper, we at first propose a novel model named Seeded Biterm Topic Model (SeedBTM) extending BTM to solve the problem of dataless short text classification with seed words. It takes advantage of both word co-occurrence information in the topic model and category-word similarity from widely used word embeddings as the prior topic-in-set knowledge. Moreover, with the same approach, we also propose Seeded Twitter Biterm Topic Model (SeedTBTM), which extends Twitter-BTM and utilizes additional user information to achieve higher classification accuracy. Experimental results on five real short-text datasets show that our models outperform the state-of-the-art methods, and especially perform well when the categories are overlapping and interrelated.
Inositol-requiring enzyme 1α (IRE1α) is the most conserved endoplasmic reticulum (ER) stress sensor with two catalytic domains, kinase and RNase, in its cytosolic portion. IRE1α inhibitors have been used to improve existing clinical treatments against various cancers. In this study we identified toxoflavin (TXF) as a new-type potent small molecule IRE1α inhibitor. We used luciferase reporter systems to screen compounds that inhibited the IRE1α-XBP1s signaling pathway. As a result, TXF was found to be the most potent IRE1α RNase inhibitor with an IC 50 value of 0.226 μM. Its inhibitory potencies on IRE1α kinase and RNase were confirmed in a series of cellular and in vitro biochemical assays. Kinetic analysis showed that TXF caused time-and reducing reagent-dependent irreversible inhibition on IRE1α, implying that ROS might participate in the inhibition process. ROS scavengers decreased the inhibition of IRE1α by TXF, confirming that ROS mediated the inhibition process. Mass spectrometry analysis revealed that the thiol groups of four conserved cysteine residues (CYS-605, CYS-630, CYS-715 and CYS-951) in IRE1α were oxidized to sulfonic groups by ROS. In molecular docking experiments we affirmed the binding of TXF with IRE1α, and predicted its binding site, suggesting that the structure of TXF itself participates in the inhibition of IRE1α. Interestingly, CYS-951 was just near the docked site. In addition, the RNase IC 50 and ROS production in vitro induced by TXF and its derivatives were negative correlated (r = −0.872). In conclusion, this study discovers a new type of IRE1α inhibitor that targets a predicted new alternative site located in the junction between RNase domain and kinase domain, and oxidizes conserved cysteine residues of IRE1α active sites to inhibit IRE1α. TXF could be used as a small molecule tool to study IRE1α's role in ER stress.
Checkpoint kinase 1 (CHK1) is a central component in DNA damage response and has emerged as a target for antitumor therapeutics. Herein, we describe the design, synthesis, and biological evaluation of a novel series of potent diaminopyrimidine CHK1 inhibitors. The compounds exhibited moderate to potent CHK1 inhibition and could suppress the proliferation of malignant hematological cell lines. The optimized compound 13 had a CHK1 IC50 value of 7.73±0.74 nM, and MV‐4‐11 cells were sensitive to it (IC50=0.035±0.007 μM). Furthermore, compound 13 was metabolically stable in mouse liver microsomes in vitro and displayed moderate oral bioavailability in vivo. Moreover, treatment of MV‐4‐11 cells with compound 13 for 2 h led to robust inhibition of CHK1 autophosphorylation on serine 296. Based on these biochemical results, we consider compound 13 to be a promising CHK1 inhibitor and potential anticancer therapeutic agent.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.