UGDAS: Unsupervised graph-network based denoiser for abstractive summarization in biomedical domain.

Abstractive summarization models can generate summary auto-regressively, but the quality is often impacted by the noise in the text. Learning cross-sentence relations is a crucial step in this task and the graph-based network is more effective to capture the sentence relationship. Moreover, knowledge is very important to distinguish the noise of the text in special domain. A novel model structure called UGDAS is proposed in this paper, which combines a sentence-level denoiser based on an unsupervised graph-network and an auto-regressive generator. It utilizes domain knowledge and sentence position information to denoise the original text and further improve the quality of generated summaries. We use the recently-introduced dataset CORD-19 (COVID-19 Open Research Dataset) on text summarization task, which contains large-scale data on coronaviruses. The experimental results show that our model achieves the SOTA (state-of-the-art) result on CORD-19 dataset and outperforms the related baseline models on the PubMed Abstract dataset.

A smartphone biosensor based on analysing structural colour of porous silicon.

We report a smartphone compatible, low-cost porous silicon biosensor, which correlates the structural colour of a porous silicon microcavity (PSiM) to spectral peak position. Molecules captured in the PSiM cause a colour change that can be quantified through image analysis. Minimal external accessories are employed. Spectrometer measurements of the PSiM reflectance spectrum shifts are carried out concurrently with the smartphone measurements to benchmark the accuracy of the smartphone biosensor. We estimate that the smartphone biosensor supports an equivalent accuracy of 0.33 nm for the detection of colour changes corresponding to spectral shifts of the PSiM. Biosensing functionality is demonstrated using a biotin-streptavidin assay with an estimated detection limit of 500 nM. The PSiM-smartphone biosensor is a promising platform for label-free point of care diagnostics.

Flow-Through Porous Silicon Membranes for Real-Time Label-Free Biosensing.

A flow-through sensing platform based on open-ended porous silicon (PSi) microcavity membranes that are compatible with integration in on-chip sensor arrays is demonstrated. Because of the high aspect ratio of PSi nanopores, the performance of closed-ended PSi sensors is limited by infiltration challenges and slow sensor responses when detecting large molecules such as proteins and nucleic acids. In order to improve molecule transport efficiency and reduce sensor response time, open-ended PSi nanopore membranes were used in a flow-through sensing scheme, allowing analyte solutions to pass through the nanopores. The molecular binding kinetics in these PSi membranes were compared through experiments and simulation with those from closed-ended PSi films of comparable thickness in a conventional flow-over sensing scheme. The flow-through PSi membrane resulted in a 6-fold improvement in sensor response time when detecting a high molecular weight analyte (streptavidin) versus in the flow-over PSi approach. This work demonstrates the possibility of integrating multiple flow-through PSi sensor membranes within parallel microarrays for rapid and multiplexed label-free biosensing.

Experimental and theoretical charge density distribution in Pigment Yellow 101.

The charge density distribution in 2,2'-dihydroxy-1,1'-naphthalazine (Pigment Yellow 101; P.Y.101) has been determined using high-resolution X-ray diffraction and multipole refinement, along with density functional theory calculations. Topological analysis of the resulting densities highlights the localisation of single/double bonds in the central C=N-N=C moiety of the molecule in its ground state. The density in the N-N is examined in detail, where we show that very small differences between experiment and theory are amplified by use of the Laplacian of the density. Quantification of hydrogen bonds highlights the importance of the intramolecular N-H···O interaction, known to be vital for retention of fluorescence in the solid state, relative to the many but weak intermolecular contacts located. However, a popular method for deriving H-bond strengths from density data appears to struggle with the intramolecular N-H···O interaction. We also show that theoretical estimation of anisotropic displacements for hydrogen atoms brings little benefit overall, and degrades agreement with experiment for one intra-molecular contact.

Correction: Experimental and theoretical charge density distribution in Pigment Yellow 101.

Correction for 'Experimental and theoretical charge density distribution in Pigment Yellow 101' by Jonathan J. Du et al., Phys. Chem. Chem. Phys., 2015, DOI: 10.1039/c4cp04302b.

In situ synthesis of peptide nucleic acids in porous silicon for drug delivery and biosensing.

Peptide nucleic acids (PNA) are a unique class of synthetic molecules that have a peptide backbone and can hybridize with nucleic acids. Here, a versatile method has been developed for the automated, in situ synthesis of PNA from a porous silicon (PSi) substrate for applications in gene therapy and biosensing. Nondestructive optical measurements were performed to monitor single base additions of PNA initiated from (3-aminopropyl)triethoxysilane attached to the surface of PSi films, and mass spectrometry was conducted to verify synthesis of the desired sequence. Comparison of in situ synthesis to postsynthesis surface conjugation of the full PNA molecules showed that surface mediated, in situ PNA synthesis increased loading 8-fold. For therapeutic proof-of-concept, controlled PNA release from PSi films was characterized in phosphate buffered saline, and PSi nanoparticles fabricated from PSi films containing in situ grown PNA complementary to micro-RNA (miR) 122 generated significant anti-miR activity in a Huh7 psiCHECK-miR122 cell line. The applicability of this platform for biosensing was also demonstrated using optical measurements that indicated selective hybridization of complementary DNA target molecules to PNA synthesized in situ on PSi films. These collective data confirm that we have established a novel PNA-PSi platform with broad utility in drug delivery and biosensing.

[A method of estimating lag between brain areas based on windowed harmonic wavelet transform].

Aiming at local field potential, the present paper introduces a method of estimating lag of neuron activities between brain areas based on windowed Harmonic wavelet transform (WHWT). Firstly, the WHWT of signals of two brain areas are calculated. Secondly, the instantaneous amplitude of the signals is calculated and finally, these amplitudes are cross-correlated and the lag at which the cross-correlation peak occurs is determined as the lag of neurons activities. Comparing with amplitude cross-correlation based on Gabor wavelet transform (GWT) or Hilbert transform (HT), this method is more precise and efficient in estimating the directionality and lag.

Improving Biomedical Question Answering by Data Augmentation and Model Weighting.

Biomedical Question Answering aims to extract an answer to the given question from a biomedical context. Due to the strong professionalism of specific domain, it's more difficult to build large-scale datasets for specific domain question answering. Existing methods are limited by the lack of training data, and the performance is not as good as in open-domain settings, especially degrading when facing to the adversarial sample. We try to resolve the above issues. First, effective data augmentation strategies are adopted to improve the model training, including slide window, summarization and round-trip translation. Second, we propose a model weighting strategy for the final answer prediction in biomedical domain, which combines the advantage of two models, open-domain model QANet and BioBERT pre-trained in biomedical domain data. Finally, we give adversarial training to reinforce the robustness of the model. The public biomedical dataset collected from PubMed provided by BioASQ challenge is used to evaluate our approach. The results show that the model performance has been improved significantly compared to the single model and other models participated in BioASQ challenge. It can learn richer semantic expression from data augmentation and adversarial samples, which is beneficial to solve more complex question answering problems in biomedical domain.

De novo pyrimidine biosynthetic complexes support cancer cell proliferation and ferroptosis defence.

De novo pyrimidine biosynthesis is achieved by cytosolic carbamoyl-phosphate synthetase II, aspartate transcarbamylase and dihydroorotase (CAD) and uridine 5'-monophosphate synthase (UMPS), and mitochondrial dihydroorotate dehydrogenase (DHODH). However, how these enzymes are orchestrated remains enigmatical. Here we show that cytosolic glutamate oxaloacetate transaminase 1 clusters with CAD and UMPS, and this complex then connects with DHODH, which is mediated by the mitochondrial outer membrane protein voltage-dependent anion-selective channel protein 3. Therefore, these proteins form a multi-enzyme complex, named 'pyrimidinosome', involving AMP-activated protein kinase (AMPK) as a regulator. Activated AMPK dissociates from the complex to enhance pyrimidinosome assembly but inactivated UMPS, which promotes DHODH-mediated ferroptosis defence. Meanwhile, cancer cells with lower expression of AMPK are more reliant on pyrimidinosome-mediated UMP biosynthesis and more vulnerable to its inhibition. Our findings reveal the role of pyrimidinosome in regulating pyrimidine flux and ferroptosis, and suggest a pharmaceutical strategy of targeting pyrimidinosome in cancer treatment.

Diethyl 4-(4-acetamido-phenyl)-2,6-dimethyl-1,4-dihydro-pyridine-3,5-dicarboxyl-ate.

The title compound, C(21)H(26)N(2)O(5), was unexpectedly obtained as a by-product in the reaction of ethyl acetoacetate, 4-acetamido-benzaldehyde and urea under microwave irradiation. The dihydro-pyridine ring assumes a flattened boat conformation. Inter-molecular N-H⋯O and weak C-H⋯O hydrogen bonding occurs in the crystal.

[Changes of local field potentials in M1 underlying the specific behavior in rat].

The local field potentials (LFPs) underlying specific behavior were recorded and analyzed in this paper from primary motor cortex (M1) with several medium, such as the self-made single channel micro-electrodes, the system of multi-channels physiological signal acquisition and processing and so on. During the experiment, the specific behavior was divided into four periods according to the changes of the recorded LFPs and the changes of the specific behavior recorded simultaneously in rats. The four periods were named prophase of catching period, planning period, catching period and the completion period, respectively. Then several methods were used for the analysis of the LFPs by MATLAB, such as time domain analysis, power spectral distribution analysis and time-frequency analysis. The results suggested that the LFPs which were caused by different behavior from a large number of movement-related neurons of M1 during the specific behavior in the process of catching play an important part in the "code" guiding role in rats. The results demonstrat that the LFPs of M1 may provide a feasibility to discriminate the motor behavior of forelimb.

Identification of purine biosynthesis as an NADH-sensing pathway to mediate energy stress.

An enhanced NADH/NAD+ ratio, termed reductive stress, is associated with many diseases. However, whether a downstream sensing pathway exists to mediate pathogenic outcomes remains unclear. Here, we generate a soluble pyridine nucleotide transhydrogenase from Escherichia coli (EcSTH), which can elevate the NADH/NAD+ ratio and meantime reduce the NADPH/NADP+ ratio. Additionally, we fuse EcSTH with previously described LbNOX (a water-forming NADH oxidase from Lactobacillus brevis) to resume the NADH/NAD+ ratio. With these tools and by using genome-wide CRISPR/Cas9 library screens and metabolic profiling in mammalian cells, we find that accumulated NADH deregulates PRPS2 (Ribose-phosphate pyrophosphokinase 2)-mediated downstream purine biosynthesis to provoke massive energy consumption, and therefore, the induction of energy stress. Blocking purine biosynthesis prevents NADH accumulation-associated cell death in vitro and tissue injury in vivo. These results underscore the pathophysiological role of deregulated purine biosynthesis in NADH accumulation-associated disorders and demonstrate the utility of EcSTH in manipulating NADH/NAD+ and NADPH/NADP+.

4-(6-Chloro-imidazo[1,2-b]pyridazin-3-yl)benzonitrile.

In the title compound, C(13)H(7)ClN(4), the imidazopyridazine ring system is essentially planar [maximum deviation 0.015 (1) Å]. It is inclined to the benzene ring of the benzonitrile group by 11.31 (2)°. In the crystal, molecules are linked via C-H⋯Cl and C-H⋯N interactions.

(1E,4Z,6E)-5-Hy-droxy-1,7-bis-(2-meth-oxy-phen-yl)-1,4,6-hepta-trien-3-one.

In the title compound, C(21)H(20)O(4), the central hepta-trienone unit is approximately planar, with a maximum atomic deviation of 0.1121 (11) Å; the two benzene rings are twisted with respect to the hepta-trienone mean plane by 2.73 (5) and 29.31 (4)°. The mol-ecule exists in the enol form and the hy-droxy group forms an intra-molecular hydrogen bond with the neighboring carbonyl group. Weak inter-molecular C-H⋯O hydrogen bonding is present in the crystal structure.

[A telemetery system for neural signal acquiring and processing].

Recording and extracting characteristic brain signals in freely moving animals is the basic and significant requirement in the study of brain-computer interface (BCI). To record animal's behaving and extract characteristic brain signals simultaneously could help understand the complex behavior of neural ensembles. Here, a system was established to record and analyse extracellular discharge in freely moving rats for the study of BCI. It comprised microelectrode and micro-driver assembly, analog front end (AFE), programmer system on chip (PSoC), wireless communication and the LabVIEW used as the platform for the graphic user interface.

Compound 968 reverses adriamycin resistance in breast cancer MCF-7ADR cells via inhibiting P-glycoprotein function independently of glutaminase.

Adriamycin (ADR) is a chemotherapeutic drug widely utilized to treat multiple types of cancers; however, the clinical efficacy of ADR is compromised due to the development of drug resistance in patients. The combination of drugs with ADR may provide a better therapeutic regimen to overcome this obstacle. Glutaminase (GLS) has been explored as a therapeutic cancer target, and its inhibition also results in increased sensitivity of tumor cells to chemotherapeutic agents. This study aimed to investigate whether GLS inhibition could reverse ADR resistance. We treated the ADR-resistant MCF-7 (MCF-7ADR) cells with a GLS inhibitor, compound 968 or CB-839, in combination with ADR. We found that compound 968, rather than CB-839, together with ADR synergistically inhibited the cell viability. These results indicated that compound 968 reversed ADR resistance in MCF-7ADR cells independently of GLS. Moreover, we modified the structure of compound 968 and finally obtained a compound 968 derivative, SY-1320, which was more potent than compound 968 in eliminating the drug resistance in MCF-7ADR cells. Furthermore, using drug affinity responsive target stability and streptavidin-biotin immunoprecipitation assays, we demonstrated that SY-1320 could specifically target P-glycoprotein (P-gp) and increase ADR accumulation through inhibition of P-gp, thereby resulting in cell death in MCF-7ADR cells. Together, our findings indicate that compound 968 or SY-1320 might be a promising drug for new combination chemotherapy in breast cancer to overcome the drug resistance.

[Relationship between serosal invasion types and lymph node metastasis after total gastrectomy in gastric cancer and its significance in selection of rational dissection].

OBJECTIVE: To evaluate the relationship between serosal invasion types and lymph node metastasis after total gastrectomy in gastric cancer patients, and explore its significance in planning practice rational dissection based on the serosa types of gastric cancer during surgery. METHODS: A total of 73 gastric cancer patients, who underwent total gastrectomy and lymph node dissection, were included in this study, and their clinicopathological data were analyzed. The serosa of gastric cancer was divided into five types: normal, reactive, nodular, tendonoid, and color-diffused, then they were combined into 3 groups: group 1: normal and reactive, group 2: nodular (including protruding nodular and flat nodular), and group 3: tendonoid and color-diffused. The lymph node metastasis ratios in the 3 groups were compared. The lymph nodes in each of the 3 groups were divided into 16 subgroups and the lymph node metastasis ratios of each subgroup in the 3 groups were compared and analyzed. RESULTS: The lymph node metastasis ratio of the gastric cancer with normal and reactive type serosa was 5.3% (26/492), the nodular was 37.1% (250/673), the tendonoid and color-diffused was 50.0% (486/972). The lymph node metastasis ratio of normal and reactive type groups was the lowest, that of the tendonoid and color-diffused groups was the highest, and the nodular type in between, showing a statistically significant difference (P<0.01). The results of comparing the lymph node metastasis ratios from the 1st to 16th subgroup in the 3 groups showed the same trend (P<0.05). CONCLUSION: Among all serosa types of gastric cancer, the lymph node metastasis ratio of the tendonoid and color-diffused is the highest, the normal and reactive type is the lowest, and the nodular in between. The extent of rational dissection should be carried out on the basis of serosa types of gastric cancer during surgery. An extended dissection including D2 and D3 lymphadenectomy should be performed for the patients with tendonoid and color-diffused serosa, a rational decreased operation including D1-D1+ lymphadenectomy should be performed for those with a normal and reactive type serosa, and for the patients with nodular type serosa, we suggest performing standard D2 dissection.

[Features of metastasis in different lymph node groups and their significance in lymph node dissection in total gastrectomy for gastric cancer].

OBJECTIVE: To summarize the features of metastasis in different lymph node groups (from 1 to 16 groups) in gastric cancer patients treated by total gastrectomy, and evaluate their clinical significance in lymph node dissection. METHODS: The data of 73 gastric cancer patients with total gastrectomy and lymph node dissection from January 2004 to April 2006 were analyzed retrospectively. The lymph nodes were divided into 16 groups according to the 13(th) edition of gastric cancer treatment guideline of JGCA (The Japan Gastric Cancer Association). The metastatic rate and degree of dissected lymph nodes in these patients were compared. RESULTS: The metastatic rates of lymph node groups in these patients from lower to higher were as follows: group 15, 13/16, 14v, 12, 10, 9, 11, 8, 2, 6/7, 5, 1, 4, 3. The lowest was the 15(th) group lymph nodes (1.4%), the highest was the 3rd group (65.8%), with a statistically significant difference between those two groups (P < 0.01). The metastatic degrees of the lymph node groups from lower to higher were as follows: 13, 16, 1, 7, 6, 5, 12, 4, 11, 8, 2, 15, 9, 3, 10, 14v. There was a statistically significant difference between the lowest group of lymph node (13(th) group, 10.7%) and the highest (14v(th), 56.3%, P < 0.01). CONCLUSION: In the radical total gastrectomy for patients with gastric cancer, it is suggested that the regional lymph nodes with higher metastatic rate should be resected necessarily, and the group with a higher metastatic degree should be dissected completely. If the result of sentinel lymph node biopsy in the 3(rd) or 14v(th) group is negative, the operation extent can be reduced. If positive, it should be extended. When the biopsy result in the 13(th) or 16(th) is positive, palliative operation may be indicated. However, if the biopsy result is negative in the 13(th) or 16(th), but positive in the 14v(th) group, extended operation is indicated.

Variation of the characteristics of biofilm on the semi-suspended bio-carrier produced by a 3D printing technique: Investigation of a whole growing cycle.

The presented investigation focused on exploring the characteristics of the biofilm formed on a novel semi-suspended bio-carrier and revealing their variation during the whole growing cycle. This used semi-suspended bio-carrier was designed to be a spindle shape, and then fabricated by using a 3D printing technique. Results indicated the bio-carrier provided a suitable environment for the attachment of diverse microorganisms. During the experimental period lasted for 45days, the biofilm quickly attached on the surface of the bio-carrier and grew to maturity, but its characteristics, including the chemical compositions, adhesion force, surface roughness, structure of microbial communities, varied continuously along with the operational time, which greatly influenced the performance of the bioreactor. The shape and structure of bio-carrier, and the shearing force caused by the aeration are important factors that influence the microbial community and its structure, and also heavily affect the formation and growth of biofilm.

Determination of the profile of DO and its mass transferring coefficient in a biofilm reactor packed with semi-suspended bio-carriers.

The work aims at illustrating the profile of DO and its mass transferring process in a biofilm reactor packed with a novel semi-suspended bio-carrier, and further revealing the main factors that influence the mass transferring coefficient of DO within the biofilm. Results showed that the biofilm was very easy to attach and grow on the semi-suspended bio-carrier, which obviously changed the DO profile inside and outside the biofilm. The semi-suspended bio-carrier caused three different mass transfer zones occurring in the bioreactor, including the zones of bulk solution, boundary layer and biofilm, in which, the boundary layer zone had an obvious higher mass transfer resistance. Increasing the aeration rate might improve the hydrodynamic conditions in the bioreactor and accelerate the mass transfer of DO, but it also detached the biofilm from the surface of bio-carrier, which reduced the consumption of DO, and accordingly, decreased the DO gradient in the bioreactor.