Compared with other methods, cyanoacrylate and lauromacrogol in treating esophagogastric varices did not increase the risk of postoperative infection.

BACKGROUND: Patients with cirrhosis commonly undergo endoscopic cyanoacrylate injection for gastric and esophageal variceal bleeding. However, postoperative infections can increase the risk of rebleeding and mortality. AIM: This study aimed to determine the risk of postoperative infections and its associated factors following cyanoacrylate injection treatment in these patients. METHODS: A retrospective analysis was conducted on 57 patients treated with ligation (ligation group), 66 patients treated with cyanoacrylate injection (injection group), and 91 patients treated with conservative treatment (control group) at the Nanchong Central Hospital. RESULTS: The rate of postoperative infection was similar among the cyanoacrylate, ligation, and conservative treatment groups, with no significant statistical difference observed ( P  = 0.97). Multivariate analysis identified postoperative Child-Pugh score and renal insufficiency as two independent risk factors for postoperative infection. The rebleeding rate in the injection group was significantly lower than in the other groups ( P  = 0.01). Mortality was significantly higher in the control group compared with the ligation and injection groups ( P  = 0.01). CONCLUSION: Cyanoacrylate combined with lauromacrogol injection did not significantly increase the risk of infection compared with ligation and conservative treatments, and it was more effective in reducing the risk of rebleeding. This method is safe, effective, and holds clinical value for broader application.

Vector-guided Fourier single-pixel imaging.

The Fourier single-pixel imaging technique exhibits great potential for compressive imaging. However, the utilization of low sampling ratio can introduce unwanted ringing artifacts, thereby compromising the fidelity of reconstructed image detail. To address this issue, Vector guided Fourier single-pixel imaging (V-FSI) has been proposed. We analyze the statistical properties in the edge vector field derived from images with low sampling ratio. Based on this information, a tailored sampling map is designed to acquire the significant high-frequency components for image reconstruction. Experimental results demonstrate the remarkable effectiveness of the proposed V-FSI method in enhancing image quality. Notably, V-FSI exhibits exceptional capabilities in perceiving and preserving the details of the objects, particularly for objects characterized by pronounced periodicity and directionality.

Measurement method of hot forging dimension based on green laser depth information.

The application field of large forgings is extensive. Accurate dimensional measurement is an important factor to ensure the quality of the finished forging product when it is forged at high temperature. Therefore, this paper proposes a green laser scanning measurement method based on depth information. First, a geometric measurement model based on depth information is established by studying the relationship between green laser depth information and forging dimension. Then, based on the heat transfer theory, distribution of the temperature field around hot forgings is studied, and an error function caused by light refraction is established using a ray tracing algorithm. After that, the error function is used to modify the measurement model to obtain the accurate distance between the forging and background plane and then obtain the dimensional information of the forging. Finally, this measurement method was experimentally verified in the laboratory, and the experimental results show that the measurement error of this method meets the dimensional measurement requirements of large hot forgings.

Sparse R-CNN: An End-to-End Framework for Object Detection.

Object detection serves as one of most fundamental computer vision tasks. Existing works on object detection heavily rely on dense object candidates, such as k anchor boxes pre-defined on all grids of an image feature map of size H×W. In this paper, we present Sparse R-CNN, a very simple and sparse method for object detection in images. In our method, a fixed sparse set of learned object proposals ( N in total) are provided to the object recognition head to perform classification and localization. By replacing HWk (up to hundreds of thousands) hand-designed object candidates with N (e.g., 100) learnable proposals, Sparse R-CNN makes all efforts related to object candidates design and one-to-many label assignment completely obsolete. More importantly, Sparse R-CNN directly outputs predictions without the non-maximum suppression (NMS) post-processing procedure. Thus, it establishes an end-to-end object detection framework. Sparse R-CNN demonstrates highly competitive accuracy, run-time and training convergence performance with the well-established detector baselines on the challenging COCO dataset and CrowdHuman dataset. We hope that our work can inspire re-thinking the convention of dense prior in object detectors and designing new high-performance detectors.

Self-Supervised Learning by Estimating Twin Class Distribution.

We present Twist, a simple and theoretically explainable self-supervised representation learning method by classifying large-scale unlabeled datasets in an end-to-end way. We employ a siamese network terminated by a softmax operation to produce twin class distributions of two augmented images. Without supervision, we enforce the class distributions of different augmentations to be consistent. However, simply minimizing the divergence between augmentations will generate collapsed solutions, i.e., outputting the same class distribution for all images. In this case, little information about the input images is preserved. To solve this problem, we propose to maximize the mutual information between the input image and the output class predictions. Specifically, we minimize the entropy of the distribution for each sample to make the class prediction assertive, and maximize the entropy of the mean distribution to make the predictions of different samples diverse. In this way, Twist can naturally avoid the collapsed solutions without specific designs such as asymmetric network, stop-gradient operation, or momentum encoder. As a result, Twist outperforms previous state-of-the-art methods on a wide range of tasks. Specifically on the semi-supervised classification task, Twist achieves 61.2% top-1 accuracy with 1% ImageNet labels using a ResNet-50 as backbone, surpassing previous best results by an improvement of 6.2%. Codes and pre-trained models are available at https://github.com/bytedance/TWIST.

Characteristic flavor metabolic network of fish sauce microbiota with different fermentation processes based on metagenomics.

This article purposed to discuss the connection between microbiota and characteristic flavor of different fish sauces (Natural fermentation (WQ), koji outdoor fermentation (YQ), heat preservation with enzyme (BWE), and heat preservation with koji (BWQ)) at the early (3 months) and late stage (7 months). A total of 117 flavor compounds were determined according to SPME-GC-MS analysis. O2PLS-DA and VIP values were used to reveal 15 and 28 flavor markers of different fish sauces at 3 and 7 M of fermentation. Further, the possible flavor formation pathways were analyzed using metagenomic sequencing, and the key microbes associated with flavor formation were identified at the genetic level. The top 10 genera related to flavor generation, such as Lactobacillus, Staphylococcus, Enterobacter, etc., appeared to play a prominent part in the flavor formation of fish sauce. The difference was that only BWQ and BWE groups could produce ethyl-alcohol through amino acid metabolism, while YQ, BWE and BWQ groups could generate phenylacetaldehyde through the transformation of Phe by α-ketoacid decarboxylase and aromatic amino acid transferase. Our research contributes to clarifying the various metabolic roles of microorganisms in the flavor generation of fish sauce.

Scale-Aware Automatic Augmentations for Object Detection With Dynamic Training.

Data augmentation is a critical technique in object detection, especially the augmentations targeting at scale invariance training (scale-aware augmentation). However, there has been little systematic investigation of how to design scale-aware data augmentation for object detection. We propose Scale-aware AutoAug to learn data augmentation policies for object detection. We define a new scale-aware search space, where both image- and instance-level augmentations are designed for maintaining scale robust feature learning. Upon this search space, we propose a new search metric, termed Pareto Scale Balance, to facilitate efficient augmentation policy search. In experiments, Scale-aware AutoAug yields significant and consistent improvement on various object detectors (e.g., RetinaNet, Faster R-CNN, Mask R-CNN, and FCOS), even compared with strong multi-scale training baselines. Our searched augmentation policies are generalized well to other datasets and instance-level tasks beyond object detection, e.g., instance segmentation. The search cost is much less than previous automated augmentation approaches for object detection, i.e., 8 GPUs across 2.5 days versus. 800 TPU-days. In addition, meaningful patterns can be summarized from our searched policies, which intuitively provide valuable knowledge for hand-crafted data augmentation design. Based on the searched scale-aware augmentation policies, we further introduce a dynamic training paradigm to adaptively determine specific augmentation policy usage during training. The dynamic paradigm consists of an heuristic manner for image-level augmentations and a differentiable copy-paste-based method for instance-level augmentations. The dynamic paradigm achieves further performance improvements to Scale-aware AutoAug without any additional burden on the long tailed LVIS benchmarks. We also demonstrate its ability to prevent over-fitting for large models, e.g., the Swin Transformer large model. Code and models are available at https://github.com/dvlab-research/SA-AutoAug.

Regulation of microbial metabolism on the formation of characteristic flavor and quality formation in the traditional fish sauce during fermentation: a review.

Fish sauce is a special flavored condiment formed by traditional fermentation of low-value fish in coastal areas, which are consumed and produced in many parts of the world, especially in Southeast Asia. In the process of fish sauce fermentation, the diversity of microbial flora and the complex metabolic reactions of microorganisms, especially lipid oxidation, carbohydrate fermentation and protein degradation, are accompanied by the formation of flavor substances. However, the precise reaction of microorganisms during the fersmentation process is difficult to accurately control in modern industrial production, which leads to the loss of traditional characteristic flavors in fermented fish sauces. This paper reviews the manufacturing processes, core microorganisms, metabolic characteristics and flavor formation mechanisms of fermented fish sauces at home and abroad. Various methods have been utilized to analyze and characterize the composition and function of microorganisms. Additionally, the potential safety issues of fermented fish sauces and their health benefits are also reviewed. Furthermore, some future directions and prospects of fermented fish sauces are also reviewed in this paper. By comprehensive understanding of this review, it is expected to address the challenges in the modern production of fish sauce thereby expanding its application in food or diet.

Antioxidant monoammonium glycyrrhizinate alleviates damage from oxidative stress in perinatal cows.

This study was conducted to evaluate the antioxidant capability of dietary supplementation with monoammonium glycyrrhizinate (MAG) in perinatal cows. Glycyrrhizic acid has been shown to have strong antioxidant activity and we hypothesised that the aglycone of glycyrrhizin and MAG, could reduce damage from oxidative stress in perinatal cows by enhancing antioxidant capacity. Blood and milk samples were collected from three groups of healthy perinatal cows that were similar in body weight, parity, milk yield in the last milk cycle, etc., receiving dietary MAG supplementation ([Day 0 = parturition]: 0 g/day, [n = 13)] 3 g/day [n = 13] or 6 g/day [n = 11]) from -28 to 56 day (0 day = parturition). Compared with 0 g/day controls (CON), milk fat was significantly decreased in cows fed with MAG, and 3 g/day had the greatest effect. A diet containing 3 g/day MAG decreased the serum alanine aminotransferase (ALT) level compared with CON at -7 day post-partum. ALT was also lower at 5 day post-partum in cows fed with 3 g/day MAG compared to 6 g/day. The administration of 3 g/day and 6 g/day MAG decreased serum aspartate transaminase (AST) at 3 day post-partum. Supplementation of MAG in cows increased total antioxidant capacity (T-AOC) in serum, and cows given 3 g MAG per day had higher T-AOC than controls on post-partum 7 day. At the end of the experiment, we isolated and cultured primary hepatocytes to determine the effect of MAG on oxidative stress caused by incubation with the sodium oleate (SO). SO increased lipid synthesis, but pre-treatment with MAG prevented the fatty buildup. SO treatment increased AST and ALT levels and malondialdehyde concentration, but decreased T-AOC and superoxide dismutase (SOD). Incubation with MAG increased antioxidant capacity and inhibited oxidant damage in bovine hepatocytes. SO stimulated expression of the antioxidant genes, NAD(P)H quinone dehydrogenase 1 (NQO1) and SOD1, in the nuclear factor erythroid 2-related factor 2 (NRF2) pathway, and catalase 1 (CAT1); this increase was accentuated by MAG pre-treatment. The results suggest that MAG can alleviate the damage caused by oxidative stress in perinatal cows by enhancing antioxidant activity.

Genetic risk of hyperuricemia in hypertensive patients associated with antihypertensive drug therapy: A longitudinal study.

Elevated serum uric acid (UA) level has been shown to be influenced by multiple genetic variants, but it remains uncertain how UA-associated variants differ in their influence on hyperuricemia risk in people taking antihypertensive drugs. We examined a total of 43 UA-related variants at 29 genes in 1840 patients with hypertension from a community-based longitudinal cohort during a median 2.25-year follow-up (including 1031 participants with normal UA, 440 prevalent hyperuricemia at baseline, and 369 new-onset hyperuricemia). Compared with the wild-type genotypes, patients carrying the SLC2A9 rs3775948G allele or the rs13129697G allele had decreased risk of hyperuricemia, while patients carrying the SLC2A9 rs11722228T allele had increased risk of hyperuricemia, after adjustment for cardiovascular risk factors and correction for multiple comparisons; moreover, these associations were modified by the use of diuretics, ß-blockers, or angiotensin converting enzyme inhibitors. The rs10821905A allele of A1CF gene was associated with increased risk of hyperuricemia, and this risk was enhanced by diuretics use. The studied variants were not observed to confer risk for incident cardiovascular events during the follow-up. In conclusion, the genes SLC2A9 and A1CF may serve as potential genetic markers for hyperuricemia risk in relation to antihypertensive drugs therapy in Chinese hypertensive patients.

The influence of immune stress induced by Escherichia coli lipopolysaccharide on the pharmacokinetics of danofloxacin in broilers.

This study aimed to determine whether the challenge from Escherichia coli (E. coli) lipopolysaccharide (LPS) affects the pharmacokinetics of danofloxacin in broilers. Twenty 1-day-old Arbor Acres (AA) broilers were equally and randomly divided into 2 groups. When the chickens were 23, 25, 27, and 29 days old, E. coli LPS (1 mL; 0.5 mg/kg body weight [BW]) and sterile saline (1 mL) were intraperitoneally injected into the two groups. After the last injection, danofloxacin was given to all chickens by gavage at the dose of 5 mg/kg BW. Then serum and plasma samples at each time point were collected through the wing vein. Danofloxacin concentrations in plasma were detected through the high-performance liquid chromatography (HPLC) method and subjected to noncompartmental analysis using Phoenix software. The levels of chicken interleukin-1ß (IL-1ß) and corticosterone (CORT) in serum were measured by the Enzyme-linked immunosorbent assay (ELISA) kit. In addition, after the collection of plasma or serum samples, 7 chickens (31 days of age) in each group were killed to calculate the organ indices. Compared with the control group, the challenge of LPS significantly decreased the parameters of AUC0-∞, Cmax, and t1/2λz and increased the parameters of Tmax and λz. Additionally, in the LPS group, the absorption time of danofloxacin was prolonged; however, the elimination was accelerated, which resulted in reduced internal exposure.

SOLO: A Simple Framework for Instance Segmentation.

Compared to many other dense prediction tasks, e.g., semantic segmentation, it is the arbitrary number of instances that has made instance segmentation much more challenging. In order to predict a mask for each instance, mainstream approaches either follow the "detect-then-segment" strategy (e.g., Mask R-CNN), or predict embedding vectors first then cluster pixels into individual instances. In this paper, we view the task of instance segmentation from a completely new perspective by introducing the notion of "instance categories", which assigns categories to each pixel within an instance according to the instance's location. With this notion, we propose segmenting objects by locations (SOLO), a simple, direct, and fast framework for instance segmentation with strong performance. We derive a few SOLO variants (e.g., Vanilla SOLO, Decoupled SOLO, Dynamic SOLO) following the basic principle. Our method directly maps a raw input image to the desired object categories and instance masks, eliminating the need for the grouping post-processing or the bounding box detection. Our approach achieves state-of-the-art results for instance segmentation in terms of both speed and accuracy, while being considerably simpler than the existing methods. Besides instance segmentation, our method yields state-of-the-art results in object detection (from our mask byproduct) and panoptic segmentation. We further demonstrate the flexibility and high-quality segmentation of SOLO by extending it to perform one-stage instance-level image matting. Code is available at: https://git.io/AdelaiDet.

Evaluation of blood pressure lowering effect by generic and brand-name antihypertensive drugs treatment: a multicenter prospective study in China.

BACKGROUND: Generic drugs are bioequivalent to their brand-name counterparts; however, concerns still exist regarding the effectiveness and safety of generic drugs because of small sample sizes and short follow-up time in most studies. The purpose of this study was to evaluate the long-term antihypertensive efficacy, cost-effectiveness and cardiovascular outcomes of generic drugs compared with brand-name drugs. METHODS: In a multicenter, community-based study including 7955 hypertensive patients who were prospectively followed up for an average of 2.5 years, we used the propensity-score-matching method to match the patients using brand-name drugs to those using generic drugs in a ratio of 1:2, 2176 patients using brand-name drugs and 4352 patients using generic drugs. RESULTS: There were no significant differences between generic drugs and brand-name drugs in blood pressure (BP)-lowering efficacy, BP control rate, and cardiovascular outcomes including coronary heart disease and stroke. The adjusted mean (95% confidence interval [CI]) of systolic BP (SBP)-lowering was -7.9 mmHg (95% CI, -9.9 to -5.9) in the brand-name drug group and -7.1 mmHg (95% CI, -9.1 to -5.1) in the generic drug group after adjusting for age, sex, body mass index, number of antihypertensive drugs and traditionally cardiovascular risk factors. Among patients aged <60 years, brand-name drugs had a higher BP control rate (47% vs. 41%; P = 0.02) and a greater effect in lowering SBP compared with generic drugs, with the between-group difference of 1.5 mmHg (95% CI, 0.2-2.8; P = 0.03). BP control rate was higher in male patients using brand-name drugs compared with those using generic drugs (46% vs. 40%; P = 0.01). Generic drugs treatment yielded an average annual incremental cost-effectiveness ratio of $315.4 per patient per mmHg decrease in SBP compared with brand-name drugs treatment. CONCLUSIONS: Our data suggested that generic drugs are suitable and cost-effective in improving hypertension management and facilitating public health benefits, especially in low- and middle-income areas.

Epigallocatechin-3-gallate activates the AMP-activated protein kinase signaling pathway to reduce lipid accumulation in canine hepatocytes.

Epigallocatechin-3-gallate (EGCG) plays a crucial role in hepatic lipid metabolism. However, the underlying regulatory mechanism of hepatic lipid metabolism by EGCG in canine is unclear. Primary canine hepatocytes were treated with EGCG (0.01, 0.1, or 1 µM) and BML-275 (an AMP-activated protein kinase [AMPK] inhibitor) to study the effects of EGCG on the gene and protein expressions associated with AMPK signaling pathway. Data showed that treatment with EGCG had greater activation of AMPK, as well as greater expression levels and transcriptional activity of peroxisome proliferator activated receptor-α (PPARα) along with upregulated messenger RNA (mRNA) abundance and protein abundance of PPARα-target genes. EGCG decreased the expression levels and transcriptional activity of sterol regulatory element-binding protein 1c (SREBP-1c) along with downregulated mRNA abundance and protein abundance of SREBP-1c target genes. Of particular interest, exogenous BML-275 could reduce or eliminate the effects of EGCG on lipid metabolism in canine hepatocytes. Furthermore, the content of triglyceride was significantly decreased in the EGCG-treated groups. These results suggest that EGCG might be a potential agent in preventing high-fat diet-induced lipid accumulation in small animals.

Longitudinal Association of Telomere Attrition with the Effects of Antihypertensive Treatment and Blood Pressure Lowering.

Leukocytes telomere length has been associated with hypertension, but, whether longitudinal telomeres change could serve as a useful predictive tool in hypertension remains uncertain. This study aimed to examine the longitudinal trajectory of leukocytes telomere length in a population-based prospective study of 1,108 individuals with hypertension. Leukocytes telomere length were measured at baseline and again after a median 2.2 (range 1.5-2.4) years of follow-up. Age as an independent predictor was inversely associated with baseline telomeres and follow-up telomeres. Annual telomere attrition rate was calculated as (follow-up telomeres-baseline telomeres)/follow-up years, and participants were categorized into the shorten and the lengthen groups. Results showed that telomere lengthening was significantly correlated with decreased systolic blood pressure (SBP) (ß=-3.28; P=0.02) and pulse pressure (PP) (ß=-2.53; P=0.02), and the differences were respectively -3.3 mmHg (95%CI, -6.2 to -0.3; P=0.03) in ∆SBP and -2.4 mmHg (95%CI, -4.9 to -0.1; P=0.04) in ∆PP between two groups after adjustment for vascular risk factors and baseline blood pressures. When stratified by age and gender, the correlations were observed in women and patients ≤60 years. Furthermore, among patients using calcium channel blocker (CCB) and angiotensin receptor blocker (ARB), those with telomeres lengthening showed a significantly lower level of ∆SBP and ∆PP. There was no correlation between telomere attrition and incidence of cardiovascular events. Our data indicated that increased telomere length of leukocytes was associated with decreased SBP and PP, particularly for patients who received CCB and ARB, supporting that telomere attrition may provide new sight in clinical intervention for hypertension.

Correction to: The Effects of 50 nm Unmodified Nano-ZnO on Lipid Metabolism and Semen Quality in Male Mice.

The original version of this article unfortunately contained a mistake. The correct title should be "The Effects of 50 nm Unmodified Nano-ZnO on Lipid Metabolism and Semen Quality in Male Mice". The original article has been corrected.

Prevalence and factors associated with fast resting heart rate in hypertensive and normotensive patients.

Background: Fast resting heart rate (RHR) is easily neglected in clinical practice of hypertension treatment.Aims: We aimed to investigate the prevalence of fast RHR and associated factors in hypertensive and normotensive individuals.Methods: We retrospectively analyzed data from two cross-sectional studies conducted in China. A total of 6763 hypertensive patients and 2807 age and sex-matched normotensive subjects with complete data on resting electrocardiogram and medical history were included. Fast RHR was defined as RHR > 85 bpm.Results: The prevalence of fast RHR was higher in hypertensive patients as compared with the normotensives (14.4% vs 7.1%, P < 0.01). In both hypertensive and normotensive subjects, fast RHR appeared as a "U-type" distribution as aging and a "inverted J type" trend as body mass index (BMI) increasing. Multivariate regression analysis showed that fast RHR was associated with age >65 or <25 years old (OR = 1.32, 95% CI 1.08-1.61), BMI <18.5 kg/m2 (OR = 2.94, 95%CI 1.47-5.87) and hypercholesterolemia (OR = 1.30, 95%CI 1.10-1.53) in hypertensive patients. Fast RHR in the normotensives was associated with female (OR = 1.78, 95%CI 1.27-2.48), pre-hypertensive state (OR = 2.38, 95%CI 1.61-3.52), and rural area origin (OR = 1.50, 95%CI 1.01-2.42). Stroke and diabetes conferred closer relevance to fast RHR in both hypertensive (OR = 1.31, 95%CI 1.02-1.69 and OR 2.26, 95%CI 1.60-3.21) and normotensive individuals (OR = 2.67, 95%CI 1.36-5.21 and OR = 2.77, 95%CI 1.47-5.23).Conclusion: Fast RHR might be common in patients with hypertension. Prior stroke and diabetes history is common associated with fast RHR. Other factors associated with fast RHR seem to be different between hypertensive patients and normotensive subjects.Abbreviations: BMI: body mass index; CI: Confidence Interval; DBP: diastolic blood pressure; ECG: electrocardiogram; OR: odd ratio; RHR: resting heart rate; SBP: systolic blood pressure.

The Effects of 50 nm Unmodified Nano-ZnO on Lipid Metabolism and Semen Quality in Male Mice.

Fifty male mice were exposed to 50 nm unmodified nano-ZnO through intragastric administration for 90 days to detect the long-term effects of unmodified nano-ZnO in mice. Results showed that the blood glucose, serum follicle stimulating hormone, luteinizing hormone, testosterone, and estradiol were significantly decreased (p < 0.05). The serum triglyceride, total cholesterol, and low-density lipoprotein were significantly increased (p < 0.05). The semen quality of the 160 mg/kg·bw group were significantly lowered (p < 0.05). The liver and testis catalase and CuZn-SOD activities were significantly elevated (p < 0.05). The abilities of •OH inhibition in the livers and testes of the 160 mg/kg·bw group were significantly lowered (p < 0.05). The liver and testis MDA levels of the 160 mg/kg·bw group were significantly elevated (p < 0.05). Results indicate that exposure of nano-ZnO could induce lipid metabolism disorder, hyperlipidemia, and reproductive toxicity to male mice through oxidative injury.

Development and Validation of Quantitative Real-Time PCR for the Detection of Residual CHO Host Cell DNA and Optimization of Sample Pretreatment Method in Biopharmaceutical Products.

BACKGROUND: The presence of residual DNA carried by biological products in the body may lead to an increased oncogenicity, infectivity, and immunomodulatory risk. Therefore, current agencies including WHO, EU, and the FDA limited the accepted amounts of residual DNA (less than 10 ng or 100 pg/dose). Among the methods of detecting residual DNA, qPCR is considered to be the most practical for residual DNA quantitation due to its sensitivity, accuracy, precision, and time-saving. RESULTS: In this study, the detection capacity of this method was determined by comparing the detected concentration of the commercial kit and the self-designed primer/probe set after the same treatment of the extraction method. Then, a universal sample pretreatment method based on a co-precipitant was optimized. The validation results demonstrated that the method has appropriate specificity, sensitivity, accuracy, and precision according to ICH guidelines. The limit of detection and quantitation reached 3 fg/ul and 0.3 pg/reaction respectively, which satisfies the requirement of limit of residual DNA detection in biologics. Spike recovery (82.3-105.7%) showed that the proposed qPCR assay was accurate and has good extraction efficiency. Moreover, the precision of the method based on intra- and inter-assay was 0.065-0.452% and 0.471-1.312%, respectively. CONCLUSIONS: These results all indicated that the method for determination of residual DNA in biological products expressed from CHO cells is sensitive, accurate and robust.