Nanjing University

About: Nanjing University is a education organization based out in Nanjing, China. It is known for research contribution in the topics: Catalysis & Population. The organization has 85961 authors who have published 105504 publications receiving 2289036 citations. The organization is also known as: NJU & Nanking University.

Multi-instance learning by treating instances as non-I.I.D. samples

Abstract: Previous studies on multi-instance learning typically treated instances in the bags as independently and identically distributed. The instances in a bag, however, are rarely independent in real tasks, and a better performance can be expected if the instances are treated in an non-i.i.d. way that exploits relations among instances. In this paper, we propose two simple yet effective methods. In the first method, we explicitly map every bag to an undirected graph and design a graph kernel for distinguishing the positive and negative bags. In the second method, we implicitly construct graphs by deriving affinity matrices and propose an efficient graph kernel considering the clique information. The effectiveness of the proposed methods are validated by experiments.

Testing and validating machine learning classifiers by metamorphic testing

Abstract: Machine learning algorithms have provided core functionality to many application domains - such as bioinformatics, computational linguistics, etc. However, it is difficult to detect faults in such applications because often there is no ''test oracle'' to verify the correctness of the computed outputs. To help address the software quality, in this paper we present a technique for testing the implementations of machine learning classification algorithms which support such applications. Our approach is based on the technique ''metamorphic testing'', which has been shown to be effective to alleviate the oracle problem. Also presented include a case study on a real-world machine learning application framework, and a discussion of how programmers implementing machine learning algorithms can avoid the common pitfalls discovered in our study. We also conduct mutation analysis and cross-validation, which reveal that our method has high effectiveness in killing mutants, and that observing expected cross-validation result alone is not sufficiently effective to detect faults in a supervised classification program. The effectiveness of metamorphic testing is further confirmed by the detection of real faults in a popular open-source classification program.

STAT3 inhibition of gluconeogenesis is downregulated by SirT1

Abstract: The fasting-activated longevity protein sirtuin 1 (SirT1, ref. 1) promotes gluconeogenesis in part, by increasing transcription of the key gluconeogenic genes pepck1 and g6pase, through deacetylating PGC-1alpha and FOXO1 (ref. 4). In contrast, signal transducer and activator of transcription 3 (STAT3) inhibits glucose production by suppressing expression of these genes. It is not known whether the inhibition of gluconeogenesis by STAT3 is controlled by metabolic regulation. Here we show that STAT3 phosphorylation and function in the liver were tightly regulated by the nutritional status of an animal, through SirT1-mediated deacetylation of key STAT3 lysine sites. The importance of the SirT1-STAT3 pathway in the regulation of gluconeogenesis was verified in STAT3-deficient mice in which the dynamic regulation of gluconeogenic genes by nutritional status was disrupted. Our results reveal a new nutrient sensing pathway through which SirT1 suppresses the inhibitory effect of STAT3, while activating the stimulatory effect of PGC-1alpha and FOXO1 on gluconeogenesis, thus ensuring maximal activation of gluconeogenic gene transcription. The connection between acetylation and phosphorylation of STAT3 implies that STAT3 may have an important role in other cellular processes that involve SirT1.

Multistimuli-Responsive Supramolecular Vesicles Based on Water-Soluble Pillar[6]arene and SAINT Complexation for Controllable Drug Release

Abstract: Supramolecular binary vesicles based on the host-guest complexation of water-soluble pillar[6]arene (WP6) and SAINT molecule have been successfully constructed, which showed pH-, Ca(2+)-, and thermal-responsiveness. These supramolecular vesicles can efficiently encapsulate model substrate calcein, which then can be efficiently released either by adjusting the solution pH to acidic condition due to the complete disruption of vesicular structure, or particularly, by adding a certain amount of Ca(2+) due to the Ca(2+)-induced vesicle fusion and accompanied by the structure disruption. More importantly, drug loading and releasing experiments demonstrate that an anticancer drug, DOX, can be successfully encapsulated by the supramolecular vesicles, and the resulting DOX-loaded vesicles exhibit efficient release of the encapsulated DOX with the pH adjustment or the introduction of Ca(2+). Cytotoxicity experiments suggest that the resulting DOX-loaded supramolecular vesicles exhibit comparable therapeutic effect for cancer cells as free DOX and the remarkably reduced damage for normal cells as well. The present multistimuli-responsive supramolecular vesicles have great potential applications in the field of controlled drug delivery. In addition, giant supramolecular vesicles (~3 μm) with large internal volume and good stability can be achieved by increasing the temperature of WP6 ⊃ SAINT vesicular solution, and they might have potential applications for bioimaging.