Identification of diabetic retinopathy classification using machine learning algorithms on clinical data and optical coherence tomography angiography.

BACKGROUND: To apply machine learning (ML) algorithms to perform multiclass diabetic retinopathy (DR) classification using both clinical data and optical coherence tomography angiography (OCTA). METHODS: In this cross-sectional observational study, clinical data and OCTA parameters from 203 diabetic patients (203 eye) were used to establish the ML models, and those from 169 diabetic patients (169 eye) were used for independent external validation. The random forest, gradient boosting machine (GBM), deep learning and logistic regression algorithms were used to identify the presence of DR, referable DR (RDR) and vision-threatening DR (VTDR). Four different variable patterns based on clinical data and OCTA variables were examined. The algorithms' performance were evaluated using receiver operating characteristic curves and the area under the curve (AUC) was used to assess predictive accuracy. RESULTS: The random forest algorithm on OCTA+clinical data-based variables and OCTA+non-laboratory factor-based variables provided the higher AUC values for DR, RDR and VTDR. The GBM algorithm produced similar results, albeit with slightly lower AUC values. Leading predictors of DR status included vessel density, retinal thickness and GCC thickness, as well as the body mass index, waist-to-hip ratio and glucose-lowering treatment. CONCLUSIONS: ML-based multiclass DR classification using OCTA and clinical data can provide reliable assistance for screening, referral, and management DR populations.

Jagged1-Notch1 Signaling Pathway Induces M1 Microglia to Disrupt the Barrier Function of Retinal Microvascular Endothelial Cells.

PURPOSE: Microglia-related inflammation is closely linked to the pathogenesis of retinal diseases. The primary objective of this research was to investigate the impact and mechanism of M1 phenotype microglia on the barrier function of retina microvascular endothelial cells. METHODS: Quantitative polymerase chain reactions and western blot techniques were utilized to analysis the mRNA and protein expressions of M1 and M2 markers of human microglial clone 3 cell line (HMC3), as well as the levels of Notch ligands and receptors under the intervention of lipopolysaccharide (LPS) or interleukin (IL)-4. ELISA was utilized to detect the pro-inflammatory and anti-inflammatory cytokines from HMC3 cells. The cellular tight junction and apoptosis of human retinal microvascular endothelial cells (HRMECs) were assessed by western blot and fluorescein isothiocyanate-dextran permeability assay. The inhibitors of Notch1 and RNA interference (RNAi) targeting Jagged1 were used to assess their contribution to the barrier function of vascular endothelial cells. RESULTS: Inducible nitric oxide synthase (iNOS) and IL-1ß were considerably elevated in LPS-treated HMC3, while CD206 and Arg-1 markedly elevated under IL-4 stimulation. The conditioned medium derived from LPS-treated HMC3 cells promoted permeability, diminished the expression of zonula occludens-1 and Occludin, and elevated the expression of Cleaved caspase-3 in HRMECs. RNAi targeting Jagged1 or Notch1 inhibitor could block M1 HMC3 polarization and maintain barrier function of HRMECs. CONCLUSION: Our findings suggest that Jagged1-Notch1 signaling pathway induces M1 microglial cells to disrupt the barrier function of HRMECs, which may lead to retinal diseases.

Dual-Stream Fusion Network with ConvNeXtV2 for Pig Weight Estimation Using RGB-D Data in Aisles.

In the field of livestock management, noncontact pig weight estimation has advanced considerably with the integration of computer vision and sensor technologies. However, real-world agricultural settings present substantial challenges for these estimation techniques, including the impacts of variable lighting and the complexities of measuring pigs in constant motion. To address these issues, we have developed an innovative algorithm, the moving pig weight estimate algorithm based on deep vision (MPWEADV). This algorithm effectively utilizes RGB and depth images to accurately estimate the weight of pigs on the move. The MPWEADV employs the advanced ConvNeXtV2 network for robust feature extraction and integrates a cutting-edge feature fusion module. Supported by a confidence map estimator, this module effectively merges information from both RGB and depth modalities, enhancing the algorithm's accuracy in determining pig weight. To demonstrate its efficacy, the MPWEADV achieved a root-mean-square error (RMSE) of 4.082 kg and a mean absolute percentage error (MAPE) of 2.383% in our test set. Comparative analyses with models replicating the latest research show the potential of the MPWEADV in unconstrained pig weight estimation practices. Our approach enables real-time assessment of pig conditions, offering valuable data support for grading and adjusting breeding plans, and holds broad prospects for application.

Millimeter-wave noise generation based on a monolithically integrated dual-mode chaotic laser.

A millimeter-wave noise generation scheme is proposed in this paper. The scheme is based on a monolithically integrated dual-mode chaotic laser, which consists of a distributed Bragg feedback (DFB) section, a phase section, and an optical amplification section. The output spectrum state of the dual-mode laser can be controlled by adjusting the injection current in the three regions. The monolithically integrated dual-mode chaotic laser has stable chaotic output and can be used as a light source for integrated millimeter-wave noise source. As a feasibility demonstration, a dual-mode chaotic laser with a mode interval of 2.05 nm was generated in the experiment, the optical mixing on a photodetector produced millimeter-wave noise with a center frequency of 259 GHz and a bandwidth of 44 GHz (237-281 GHz), achieving a typical value of excess noise ratio of 47 dB. It has the advantages of high noise source utilization, small noise source volume, and high integration.

Surface-tethered ROS-responsive micelle backpacks for boosting mesenchymal stem cell vitality and modulating inflammation in ischemic stroke treatment.

Mesenchymal stem cells (MSCs) exhibited remarkable therapeutic potential in ischemic stroke due to their exceptional immunomodulatory ability and paracrine effect; they have also been regarded as excellent neuroprotectant delivery vehicles with inflammatory tropism. However, the presence of high levels of reactive oxygen species (ROS) and an oxidative stress environment at the lesion site inhibits cell survival and further therapeutic effects. Using bioorthogonal click chemistry, ROS-responsive luteolin-loaded micelles were tethered to the surface of MSCs. As MSCs migrated to the ischemic brain, the micelles would achieve ROS-responsive release of luteolin to protect MSCs from excessive oxidative damage while inhibiting neuroinflammation and scavenging ROS to ameliorate ischemic stroke. This study provided an effective and prospective therapeutic strategy for ischemic stroke and a framework for a stem cell-based therapeutic system to treat inflammatory cerebral diseases.

Associations between socioeconomic status and stroke in American adults: A population-based study.

Stroke is an acute cerebrovascular disease that can lead to disability and death. This study aimed to investigate the relationship between socioeconomic status (SES) and stroke. SES was evaluated by two variables: poverty to income ratio (PIR) and education level. In this multi-subject study, we collected data from the National Health and Nutrition Examination Survey (NHANES) database between 2009 and 2018, and finally 22,792 adults (≥20 years old) were included in the study. We proceeded with weighted multivariate logistic regression analysis as well as subgroup analysis. When analyzing the effect of PIR on stroke alone, the results showed that an increase in PIR levels was associated with a decrease in stroke incidence (OR = 0.764 95% CI: (0.711, 0.820), p < 0.001). The multivariate analysis presented a decline in stroke incidence in the highest quartile PIR group compared to the lowest quartile PIR group (OR = 0.296 95% CI: (0.214, 0.409), P＜0.001). Our results indicated that PIR is a protective factor for stroke, but there are exceptions in this relationship among different people. Hence, it is imperative that policymakers, healthcare providers, and clinicians take into account the inequality distribution of SES among adults while developing and executing stroke prevention and treatment strategies.

An Improved Pig Counting Algorithm Based on YOLOv5 and DeepSORT Model.

Pig counting is an important task in pig sales and breeding supervision. Currently, manual counting is low-efficiency and high-cost and presents challenges in terms of statistical analysis. In response to the difficulties faced in pig part feature detection, the loss of tracking due to rapid movement, and the large counting deviation in pig video tracking and counting research, this paper proposes an improved pig counting algorithm (Mobile Pig Counting Algorithm with YOLOv5xpig and DeepSORTPig (MPC-YD)) based on YOLOv5 + DeepSORT model. The algorithm improves the detection rate of pig body parts by adding two different sizes of SPP networks and using SoftPool instead of MaxPool operations in YOLOv5x. In addition, the algorithm includes a pig reidentification network, a pig-tracking method based on spatial state correction, and a pig counting method based on frame number judgment on the DeepSORT algorithm to improve pig tracking accuracy. Experimental analysis shows that the MPC-YD algorithm achieves an average precision of 99.24% in pig object detection and an accuracy of 85.32% in multitarget pig tracking. In the aisle environment of the slaughterhouse, the MPC-YD algorithm achieves a correlation coefficient (R2) of 98.14% in pig counting from video, and it achieves stable pig counting in a breeding environment. The algorithm has a wide range of application prospects.

Associations of concomitant retinopathy and depression with mortality in a nationally representative population.

OBJECTIVE: To evaluate the interaction effects between retinopathy and depression on mortality risks in genral population and subpopulation with diabetes. METHODS: Prospective analyses were conducted on data from the National Health and Nutrition Examination Surveys study. Associations of retinopathy, depression and their interaction with all-cause, cardiovascular disease (CVD)-specific, cancer-specific and other-specific mortality risk were estimated using Kaplan-Meier curves and multivariate Cox proportional hazards models. RESULTS: Among 5367 participants, the weighted prevalence of retinopathy and depression was 9.6 % and 7.1 %, respectively. After a follow-up period of 12.1 years, 1295 deaths (17.3 %) occurred. Retinopathy was associated with an increased risk of all-cause (hazard ratio [HR]; 95 % confidence interval [CI]) (1.47; 1.27-1.71), CVD-specific (1.87; 1.45-2.41), and other-specific (1.43; 1.14-1.79) mortality. Similar relationship was observed between depression and all-cause mortality (1.24; 1.02-1.52). Retinopathy and depression had a positive multiplicative and additive interaction effect on all-cause (Pinteraction = 0.015; relative excess risk of interaction [RERI] 1.30; 95 % CI 0.15-2.45) and CVD-specific mortality (Pinteraction = 0.042; RERI 2.65; 95 % CI -0.12-5.42). Concomitant retinopathy and depression was more markedly associated with all-cause (2.86; 1.91-4.28), CVD-specific (4.70; 2.57-8.62), and other-specific mortality risks (2.18; 1.14-4.15) compared to those without retinopathy and depression. These associations were more pronounced in the diabetic participants. CONCLUSIONS: The co-occurrence of retinopathy and depression increases the risk of all-cause and CVD-specific mortality among middle-aged and older adults in the United States, especially in population with diabetes. Focus on diabetic patients and active evaluation and intervention of retinopathy with depression may improve their quality of life and mortality outcomes.

Performance Degradation Modeling and Its Prediction Algorithm of an IGBT Gate Oxide Layer Based on a CNN-LSTM Network.

The problem of health status prediction of insulated-gate bipolar transistors (IGBTs) has gained significant attention in the field of health management of power electronic equipment. The performance degradation of the IGBT gate oxide layer is one of the most important failure modes. In view of failure mechanism analysis and the easy implementation of monitoring circuits, this paper selects the gate leakage current of an IGBT as the precursor parameter of gate oxide degradation, and uses time domain characteristic analysis, gray correlation degree, Mahalanobis distance, Kalman filter, and other methods to carry out feature selection and fusion. Finally, it obtains a health indicator, characterizing the degradation of IGBT gate oxide. The degradation prediction model of the IGBT gate oxide layer is constructed by the Convolutional Neural Network and Long Short-Term Memory (CNN-LSTM) Network, which show the highest fitting accuracy compared with Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), Support Vector Regression (SVR), Gaussian Process Regression (GPR), and CNN-LSTM models in our experiment. The extraction of the health indicator and the construction and verification of the degradation prediction model are carried out on the dataset released by the NASA-Ames Laboratory, and the average absolute error of performance degradation prediction is as low as 0.0216. These results show the feasibility of the gate leakage current as a precursor parameter of IGBT gate oxide layer failure, as well as the accuracy and reliability of the CNN-LSTM prediction model.

A Pig Mass Estimation Model Based on Deep Learning without Constraint.

The body mass of pigs is an essential indicator of their growth and health. Lately, contactless pig body mass estimation methods based on computer vision technology have gained attention thanks to their potential to improve animal welfare and ensure breeders' safety. Nonetheless, current methods require pigs to be restrained in a confinement pen, and no study has been conducted in an unconstrained environment. In this study, we develop a pig mass estimation model based on deep learning, capable of estimating body mass without constraints. Our model comprises a Mask R-CNN-based pig instance segmentation algorithm, a Keypoint R-CNN-based pig keypoint detection algorithm and an improved ResNet-based pig mass estimation algorithm that includes multi-branch convolution, depthwise convolution, and an inverted bottleneck to improve accuracy. We constructed a dataset for this study using images and body mass data from 117 pigs. Our model achieved an RMSE of 3.52 kg on the test set, which is lower than that of the pig body mass estimation algorithm with ResNet and ConvNeXt as the backbone network, and the average estimation speed was 0.339 s·frame-1 Our model can evaluate the body quality of pigs in real-time to provide data support for grading and adjusting breeding plans, and has broad application prospects.

Arsenic-Loaded Biomimetic Iron Oxide Nanoparticles for Enhanced Ferroptosis-Inducing Therapy of Hepatocellular Carcinoma.

Hepatocellular carcinoma (HCC) has a poor response to most available systemic therapies due to intrinsic or acquired resistance to apoptosis. Ferroptosis-based therapy is expected to circumvent those limitations. Therefore, the rational design of ferroptosis-based therapies with targeted delivery of ferroptosis inducers for HCC is in need. In this study, we found that arsenic trioxide (ATO) can efficiently induce ferroptosis in HCC cells, and this effect could be reversed by the iron chelator deferoxamine. On this basis, a drug delivery system was constructed to enhance the therapeutic efficacy of ATO by camouflaging ATO-loaded magnetic nanoparticles (Fe3O4) with HCC cell membranes, termed AFN@CM. After AFN@CM treatment, glutathione peroxidase 4 was strongly inhibited and intracellular lipid peroxide species were significantly increased in HCC cells. In addition, enhanced ferroptosis and tumor suppression were observed both in vitro and in vivo. The bio-safety of AFN@CM was also demonstrated by the in vivo toxicity evaluation. In addition, benefiting from the cell membrane coating, AFN@CM showed enhanced accumulation at tumor sites and achieved continuous tumor elimination in the mouse model. In conclusion, AFN@CM exhibited satisfactory therapeutic efficacy in the treatment of HCC and provided a desirable ferroptosis-based strategy for safe and reliable HCC therapeutics.

Multiple Calcium Channel Types with Unique Expression Patterns Mediate Retinal Signaling at Bipolar Cell Ribbon Synapses.

Retinal bipolar cells (BCs) compose the canonical vertical excitatory pathway that conveys photoreceptor output to inner retinal neurons. Although synaptic transmission from BC terminals is thought to rely almost exclusively on Ca2+ influx through voltage-gated Ca2+ (CaV) channels mediating L-type currents, the molecular identity of CaV channels in BCs is uncertain. Therefore, we combined molecular and functional analyses to determine the expression profiles of CaV α1, ß, and α2δ subunits in mouse rod bipolar (RB) cells, BCs from which the dynamics of synaptic transmission are relatively well-characterized. We found significant heterogeneity in CaV subunit expression within the RB population from mice of either sex, and significantly, we discovered that transmission from RB synapses was mediated by Ca2+ influx through P/Q-type (CaV2.1) and N-type (CaV2.2) conductances as well as the previously-described L-type (CaV1) and T-type (CaV3) conductances. Furthermore, we found both CaV1.3 and CaV1.4 proteins located near presynaptic ribbon-type active zones in RB axon terminals, indicating that the L-type conductance is mediated by multiple CaV1 subtypes. Similarly, CaV3 α1, ß, and α2δ subunits also appear to obey a "multisubtype" rule, i.e., we observed a combination of multiple subtypes, rather than a single subtype as previously thought, for each CaV subunit in individual cells.SIGNIFICANCE STATEMENT Bipolar cells (BCs) transmit photoreceptor output to inner retinal neurons. Although synaptic transmission from BC terminals is thought to rely almost exclusively on Ca2+ influx through L-type voltage-gated Ca2+ (CaV) channels, the molecular identity of CaV channels in BCs is uncertain. Here, we report unexpectedly high molecular diversity of CaV subunits in BCs. Transmission from rod bipolar (RB) cell synapses can be mediated by Ca2+ influx through P/Q-type (CaV2.1) and N-type (CaV2.2) conductances as well as the previously-described L-type (CaV1) and T-type (CaV3) conductances. Furthermore, CaV1, CaV3, ß, and α2δ subunits appear to obey a "multisubtype" rule, i.e., a combination of multiple subtypes for each subunit in individual cells, rather than a single subtype as previously thought.

Glutamate Transporters EAAT2 and EAAT5 Differentially Shape Synaptic Transmission from Rod Bipolar Cell Terminals.

Excitatory amino acid transporters (EAATs) control visual signal transmission in the retina by rapidly removing glutamate released from photoreceptors and bipolar cells (BCs). Although it has been reported that EAAT2 and EAAT5 are expressed at presynaptic terminals of photoreceptors and some BCs in mammals, the distinct functions of these two glutamate transporters in retinal synaptic transmission, especially at a single synapse, remain elusive. In this study, we found that EAAT2 was expressed in all BC types while coexisting with EAAT5 in rod bipolar (RB) cells and several types of cone BCs from mice of either sex. Our immunohistochemical study, together with a recently published literature (Gehlen et al., 2021), showed that EAAT2 and EAAT5 were both located in RB axon terminals near release sites. Optogenetic, electrophysiological and pharmacological analyses, however, demonstrated that EAAT2 and EAAT5 regulated neurotransmission at RB→AII amacrine cell synapses in significantly different ways: EAAT5 dramatically affected both the peak amplitude and kinetics of postsynaptic responses in AIIs, whereas EAAT2 had either relatively small or opposite effects. By contrast, blockade of EAAT1/GLAST, which was exclusively expressed in Müller cells, showed no obvious effect on AII responses, indicating that glutamate uptake by Müller cells did not influence synaptic transmission from RB terminals. Furthermore, we found that temporal resolution at RB→AII synapses was reduced substantially by blockade of EAAT5 but not EAAT2. Taken together, our work reveals the distinct functions of EAAT2 and EAAT5 in signal transmission at RB ribbon synapses.

Potential of PKM2 as a drug target in mouse models with type 1 diabetes mellitus.

BACKGROUND: This study aimed to determine the effect of PKM2 knockout in STZ induced type 1 diabetes mellitus (T1D) mouse models and to explore the possible mechanism. METHOD: PKM2fl/fl C57BL/6 mouse was backcrossed with Ins-1cre C57BL/6 mouse to generate ß-cell-specific PKM2 knockout mouse after tamoxifen administration. The expression level of PKM2 in pancreas tissues was detected by quantitative reverse-transcription polymerase chain reaction and western blot analysis. The blood glucose levels in STZ induced T1D mouse models were measured to validate the establishment of T1D models. The pathological changes of T1D mouse were examined by hematoxylin and eosin. The oxidative stress (OS) and inflammatory response in T1D mouse were determined by measuring the expression levels of malondialdehyde, superoxide dismutase, and 8-OHdG in pancreatic tissues and the serum levels of interleukin-6 and tumor necrosis factor-α. The ability to catabolize glucose was assessed through intraperitoneal glucose tolerance test and insulin tolerance test. RESULTS: ß-cell-specific PKM2 knockout was successfully achieved in PKM2fl/flcre+ mouse. T1D mouse with PKM2 knockdown had decreased blood glucose level and suppressed cell apoptosis. PKM2 knockout in T1D mouse attenuated ß cell injury. OS and inflammatory response in T1D mouse with PKM2 knockout were also suppressed compared with T1D mouse without PKM2 knockout. CONCLUSION: PKM2 knockout in T1D mouse can attenuate OS and inflammatory response as well as decrease blood glucose level, suggesting the potential of PKM2 as a drug target for T1D treatment.

Genotype-phenotype correlation and interaction of 4q25, 15q14 and MIPEP variants with myopia in southern Chinese population.

BACKGROUND/AIMS: To determine the association and interaction of genome-wide association study-reported variants for Asian populations with myopia and ocular biometric parameters in southern Chinese population. METHODS: Totally, 1462 unrelated Han Chinese subjects were recruited with complete ophthalmic examinations, including 1196 myopia and 266 control subjects. A total of nine variants were selected for TaqMan genotyping. The genetic association, joint additive effect and genotype-phenotype correlation were investigated. RESULTS: The 4q25 variant rs10034228 (p=0.002, OR=0.56) and MIPEP variant rs9318086 (p=0.004, OR=1.62) were found to be significantly associated with myopia as well as different severity of myopia. Moreover, 15q14 variant rs524952 (p=0.015, OR=1.49) also showed mild association with myopia and high myopia. However, there was no significant association of CTNND2, vasoactive intestinal peptide receptor 2 and syntrophin beta 1 variants with myopia. Joint additive analysis revealed that the subjects carrying 6 risk alleles of the 3 associated variants were 10-fold higher risk predisposed to high myopia. Genotype-phenotype correlation analysis revealed that high myopia subjects carrying 4q25 rs10034228 T allele showed thicker central corneal thickness, whereas high myopia subjects carrying 15q14 rs524952 A allele were associated with longer axial length and larger curvature ratio. CONCLUSION: This study revealed significant association of 4q25, 15q14 and MIPEP variants with myopia and different severity of myopia in southern Chinese population, joint additively enhancing 10-fold of risk predisposing to high myopia. The correlation of these associated variants with axial length and corneal parameters suggests their contribution to the refractive status in high myopia subjects.

Dual carbon decorated germanium-carbon composite as a stable anode for sodium/potassium-ion batteries.

In the present work, we introduce a dual carbon accommodated structure in which germanium nanoparticles are encapsulated into an ordered mesoporous carbon matrix (Ge-CMK) and further coated with an amorphous carbon layer (Ge@C-CMK) through a nano-casting route followed by chemical vapor deposition (CVD) treatment. In the resultant Ge@C-CMK composite, the unique lane-like pore structure that cooperates with the amorphous carbon surface can not only mitigate the volume expansion of germanium particles, but also improve the electrical conductivity of germanium as well as facilitate Na+/K+ diffusion. When employed as the anode of sodium-ion batteries, the Ge@C-CMK electrode exhibits stable capacity as well as long-term cycling stability (a stable capacity of 176 mAh g-1 at 1 A g-1 after 5000 cycles). Furthermore, it also delivers a reversible capacity when used as the anode of potassium-ion batteries. This demonstrates that the Ge@C-CMK electrode possesses promising application potential as an alternative anode in sodium and potassium ion storage applications.

Facile fabrication of a vanadium nitride/carbon fiber composite for half/full sodium-ion and potassium-ion batteries with long-term cycling performance.

Vanadium-based composite anodes have been designed for applications in alkali metal ion batteries, including lithium-ion batteries (LIBs), sodium-ion batteries (SIBs) and potassium-ion batteries (PIBs). However, the problems of inferior long-term cycling stability caused by the large volume change and dissolution of vanadium-based active materials during cycles and slow diffusion for large radii of Na+ and K+ still limit their underlying capability and need to be addressed. In the present work, we initially designed and fabricated a vanadium nitride/carbon fiber (VN/CNF) composite via a facile electrospinning method followed by the ammonization process. The obtained VN/CNF composite anode exhibited excellent half/full sodium and potassium storage performance. When used as an anode material for SIBs, it delivered a high capacity of 403 mA h g-1 at 0.1 A g-1 after 100 cycles and as large as 237 mA h g-1 at 2 A g-1 even after 4000 cycles with negligible capacity fading. More importantly, the VN/CNFs//Na3V2(PO4)3 full cell by coupling the VN/CNF composite anode with the Na3V2(PO4)3 (NVP) cathode also exhibited a desirable capacity of 257 mA h g-1 at 500 mA g-1 after 50 cycles. Besides, when further evaluated as an anode for PIBs, the VN/CNF composite anode achieved a large capacity of 266 mA h g-1 after 200 cycles at 0.1 A g-1 and maintained a stable capacity of 152 mA h g-1 at 1 A g-1 even after 1000 cycles, showing significant long-term cycling stability. This is one of the best performances of vanadium-based anode materials for SIBs and PIBs reported so far.

Regulation of B cell homeostasis by Ptpn22 contributes to type 1 diabetes in NOD mice.

PURPOSE: A coding variant in PTPN22 (C1858T) is one of the most important genetic risk factors in type 1 diabetes (T1D). The role of the PTPN22 risk allele in B cells is still incompletely understood and has not been investigated directly in T1D. This study aimed to explore the role of PTPN22 in the homeostasis of B cells and its influence in T1D. METHODS: Wild-type (WT) and Ptpn22 inducible knockdown (KD) NOD mice were treated with 200 µg/ml doxycycline at the age of 10 weeks for 1-2 months. B cell compositions in the bone marrow, peritoneal cavity and spleen were examined. The pathogenicity of Ptpn22 KD B cells was explored by adoptive cell transfer. RESULTS: Ptpn22 silencing increased the frequency of recirculating mature B cells in the bone marrow, decreased the frequency of B-1a cells in the peritoneal cavity and suppressed the formation of marginal zone B cells and plasma cells in the spleen. Changes in the composition of the peripheral B cell compartment caused by altered cell proliferation while rates of apoptosis were not affected. Significantly, co-transfer of Ptpn22 KD B cells with NY8.3 diabetogenic T cells diminished the frequency of diabetes in recipient NOD.scid mice compared with co-transfer of WT B cells. CONCLUSIONS: Our study constitutes the first functional study of Ptpn22 in B cells in NOD mice. Our findings suggest that Ptpn22 variation contributes to T1D by modifying the B cell compartment and support a gain-of-function for the PTPN22 disease variant.

Generalized regression neural network association with terahertz spectroscopy for quantitative analysis of benzoic acid additive in wheat flour.

Investigations were initiated to develop terahertz (THz) techniques associated with machine learning methods of generalized regression neural network (GRNN) and back-propagation neural network (BPNN) to rapidly measure benzoic acid (BA) content in wheat flour. The absorption coefficient exhibited a maximum absorption peak at 1.94 THz, which generally increased with the content of BA additive. THz spectra were transformed into orthogonal principal component analysis (PCA) scores as the input vectors of GRNN and BPNN models. The best GRNN model was achieved with three PCA scores and spread value of 0.2. Compared with the BPNN model, GRNN model to powder samples could be considered very successful for quality control of wheat flour with a correlation coefficient of prediction (r p) of 0.85 and root mean square error of prediction of 0.10%. The results suggest that THz technique association with GRNN has a significant potential to quantitatively analyse BA additive in wheat flour.