Rescue of a lysosomal storage disorder caused by Grn loss of function with a brain penetrant progranulin biologic.

GRN mutations cause frontotemporal dementia (GRN-FTD) due to deficiency in progranulin (PGRN), a lysosomal and secreted protein with unclear function. Here, we found that Grn-/- mice exhibit a global deficiency in bis(monoacylglycero)phosphate (BMP), an endolysosomal phospholipid we identified as a pH-dependent PGRN interactor as well as a redox-sensitive enhancer of lysosomal proteolysis and lipolysis. Grn-/- brains also showed an age-dependent, secondary storage of glucocerebrosidase substrate glucosylsphingosine. We investigated a protein replacement strategy by engineering protein transport vehicle (PTV):PGRN-a recombinant protein linking PGRN to a modified Fc domain that binds human transferrin receptor for enhanced CNS biodistribution. PTV:PGRN rescued various Grn-/- phenotypes in primary murine macrophages and human iPSC-derived microglia, including oxidative stress, lysosomal dysfunction, and endomembrane damage. Peripherally delivered PTV:PGRN corrected levels of BMP, glucosylsphingosine, and disease pathology in Grn-/- CNS, including microgliosis, lipofuscinosis, and neuronal damage. PTV:PGRN thus represents a potential biotherapeutic for GRN-FTD.

Nucleation-Inhibited Emulsion Interfacial Assembled Polydopamine Microvesicles as Artificial Antigen-Presenting Cells.

Albeit microemulsion systems have emerged as efficient platforms for fabricating tunable nano/microstructures, lack of understanding on the emulsion-interfacial assembly hindered the control of fabrication. Herein, a nucleation-inhibited microemulsion interfacial assembly method is proposed, which deviates from conventional interfacial nucleation approaches, for the synthesis of polydopamine microvesicles (PDA MVs). These PDA MVs exhibit an approximate diameter of 1 µm, showcasing a pliable structure reminiscent of cellular morphology. Through modifications of antibodies on the surface of PDA MVs, their capacity as artificial antigen presentation cells is evaluated. In comparison to solid nanoparticles, PDA MVs with cell-like structures show enhanced T-cell activation, resulting in a 1.5-fold increase in CD25 expression after 1 day and a threefold surge in PD-1 positivity after 7 days. In summary, the research elucidates the influence of nucleation and interfacial assembly in microemulsion polymerization systems, providing a direct synthesis method for MVs and substantiating their effectiveness as artificial antigen-presenting cells.

ViMRT: a text-mining tool and search engine for automated virus mutation recognition.

MOTIVATION: Virus mutation is one of the most important research issues which plays a critical role in disease progression and has prompted substantial scientific publications. Mutation extraction from published literature has become an increasingly important task, benefiting many downstream applications such as vaccine design and drug usage. However, most existing approaches have low performances in extracting virus mutation due to both lack of precise virus mutation information and their development based on human gene mutations. RESULTS: We developed ViMRT, a text-mining tool and search engine for automated virus mutation recognition using natural language processing. ViMRT mainly developed 8 optimized rules and 12 regular expressions based on a development dataset comprising 830 papers of 5 human severe disease-related viruses. It achieved higher performance than other tools in a test dataset (1662 papers, 99.17% in F1-score) and has been applied well to two other viruses, influenza virus and severe acute respiratory syndrome coronavirus-2 (212 papers, 96.99% in F1-score). These results indicate that ViMRT is a high-performance method for the extraction of virus mutation from the biomedical literature. Besides, we present a search engine for researchers to quickly find and accurately search virus mutation-related information including virus genes and related diseases. AVAILABILITY AND IMPLEMENTATION: ViMRT software is freely available at http://bmtongji.cn:1225/mutation/index.

MolSHAP: Interpreting Quantitative Structure-Activity Relationships Using Shapley Values of R-Groups.

Optimizing the activities and properties of lead compounds is an essential step in the drug discovery process. Despite recent advances in machine learning-aided drug discovery, most of the existing methods focus on making predictions for the desired objectives directly while ignoring the explanations for predictions. Although several techniques can provide interpretations for machine learning-based methods such as feature attribution, there are still gaps between these interpretations and the principles commonly adopted by medicinal chemists when designing and optimizing molecules. Here, we propose an interpretation framework, named MolSHAP, for quantitative structure-activity relationship analysis by estimating the contributions of R-groups. Instead of attributing the activities to individual input features, MolSHAP regards the R-group fragments as the basic units of interpretation, which is in accordance with the fragment-based modifications in molecule optimization. MolSHAP is a model-agnostic method that can interpret activity regression models with arbitrary input formats and model architectures. Based on the evaluations of numerous representative activity regression models on a specially designed R-group ranking task, MolSHAP achieved significantly better interpretation power compared with other methods. In addition, we developed a compound optimization algorithm based on MolSHAP and illustrated the reliability of the optimized compounds using an independent case study. These results demonstrated that MolSHAP can provide a useful tool for accurately interpreting the quantitative structure-activity relationships and rationally optimizing the compound activities in drug discovery.

Tanshinone IIA regulates CCl4 induced liver fibrosis in C57BL/6J mice via the PI3K/Akt and Nrf2/HO-1 signaling pathways.

Chronic liver diseases caused by various factors may develop into liver fibrosis (LF). Early stage of LF could be reversible. Tanshinone IIA (Tan IIA), an extract from Salvia miltiorrhiza, has been reported to be hepatoprotective. However, the potential targets and mechanism of Tan IIA in the treatment of LF are still unclear. Our study aims at the anti-LF mechanism of Tan IIA through network pharmacological analysis combined with LF-related experiments. Serum biochemical indicators and histopathological examination showed that Tan IIA could ameliorate the process of LF in the CCl4 -induced mouse model. Western blot and immunohistochemical assays showed that Tan IIA decreased the expression of Kirsten rat sarcoma viral oncogene homolog (KRAS), phosphatidylinositide 3-kinases/protein kinase B (PI3K/Akt), and nuclear factor erythroid 2-related factor/heme oxygenase-1 (Nrf2/HO-1). Compared with the model group, the Tan IIA groups increased the decreased superoxide dismutase activity and glutathione content, while decreasing the increased malondialdehyde content. These results indicate that Tan IIA may play an antioxidant role by inhibiting the expression of KRAS, PI3K/Akt, and Nrf2/HO-1 to ameliorate the progression of LF, which to some extent explains the pharmacological mechanism of Tan IIA in LF. In conclusion, our study demonstrates that Tan IIA could regulate LF via PI3K/Akt and Nrf2/HO-1 signaling pathways. It may be an effective therapeutic compound for the treatment of LF.

A machine learning-based framework for modeling transcription elongation.

RNA polymerase II (Pol II) generally pauses at certain positions along gene bodies, thereby interrupting the transcription elongation process, which is often coupled with various important biological functions, such as precursor mRNA splicing and gene expression regulation. Characterizing the transcriptional elongation dynamics can thus help us understand many essential biological processes in eukaryotic cells. However, experimentally measuring Pol II elongation rates is generally time and resource consuming. We developed PEPMAN (polymerase II elongation pausing modeling through attention-based deep neural network), a deep learning-based model that accurately predicts Pol II pausing sites based on the native elongating transcript sequencing (NET-seq) data. Through fully taking advantage of the attention mechanism, PEPMAN is able to decipher important sequence features underlying Pol II pausing. More importantly, we demonstrated that the analyses of the PEPMAN-predicted results around various types of alternative splicing sites can provide useful clues into understanding the cotranscriptional splicing events. In addition, associating the PEPMAN prediction results with different epigenetic features can help reveal important factors related to the transcription elongation process. All these results demonstrated that PEPMAN can provide a useful and effective tool for modeling transcription elongation and understanding the related biological factors from available high-throughput sequencing data.

Boris knockout eliminates AOM/DSS-induced in situ colorectal cancer by suppressing DNA damage repair and inflammation.

The Brother of Regulator of Imprinted Sites (BORIS, gene symbol CTCFL) has previously been shown to promote colorectal cancer cell proliferation, inhibit cancer cell apoptosis, and resist chemotherapy. However, it is unknown whether Boris plays a role in the progression of in situ colorectal cancer. Here Boris knockout (KO) mice were constructed. The function loss of the cloned Boris mutation that was retained in KO mice was verified by testing its activities in colorectal cell lines compared with the Boris wild-type gene. Boris knockout reduced the incidence and severity of azoxymethane/dextran sulfate-sodium (AOM/DSS)-induced colon cancer. The importance of Boris is emphasized in the progression of in situ colorectal cancer. Boris knockout significantly promoted the phosphorylation of γH2AX and the DNA damage in colorectal cancer tissues and suppressed Wnt and MAPK pathways that are responsible for the callback of DNA damage repair. This indicates the strong inhibition of colorectal cancer in Boris KO mice. By considering that the DSS-promoted inflammation contributes to tumorigenesis, Boris KO mice were also studied in DSS-induced colitis. Our data showed that Boris knockout alleviated DSS-induced colitis and that Boris knockdown inhibited the NF-κB signaling pathway in RAW264.7 cells. Therefore Boris knockout eliminates colorectal cancer generation by inhibiting DNA damage repair in cancer cells and relieving inflammation in macrophages. Our findings demonstrate the importance of Boris in the development of in situ colorectal cancer and provide evidence for the feasibility of colorectal cancer therapy on Boris.

Large-Scale Biomedical Relation Extraction Across Diverse Relation Types: Model Development and Usability Study on COVID-19.

BACKGROUND: Biomedical relation extraction (RE) is of great importance for researchers to conduct systematic biomedical studies. It not only helps knowledge mining, such as knowledge graphs and novel knowledge discovery, but also promotes translational applications, such as clinical diagnosis, decision-making, and precision medicine. However, the relations between biomedical entities are complex and diverse, and comprehensive biomedical RE is not yet well established. OBJECTIVE: We aimed to investigate and improve large-scale RE with diverse relation types and conduct usability studies with application scenarios to optimize biomedical text mining. METHODS: Data sets containing 125 relation types with different entity semantic levels were constructed to evaluate the impact of entity semantic information on RE, and performance analysis was conducted on different model architectures and domain models. This study also proposed a continued pretraining strategy and integrated models with scripts into a tool. Furthermore, this study applied RE to the COVID-19 corpus with article topics and application scenarios of clinical interest to assess and demonstrate its biological interpretability and usability. RESULTS: The performance analysis revealed that RE achieves the best performance when the detailed semantic type is provided. For a single model, PubMedBERT with continued pretraining performed the best, with an F1-score of 0.8998. Usability studies on COVID-19 demonstrated the interpretability and usability of RE, and a relation graph database was constructed, which was used to reveal existing and novel drug paths with edge explanations. The models (including pretrained and fine-tuned models), integrated tool (Docker), and generated data (including the COVID-19 relation graph database and drug paths) have been made publicly available to the biomedical text mining community and clinical researchers. CONCLUSIONS: This study provided a comprehensive analysis of RE with diverse relation types. Optimized RE models and tools for diverse relation types were developed, which can be widely used in biomedical text mining. Our usability studies provided a proof-of-concept demonstration of how large-scale RE can be leveraged to facilitate novel research.

Recent advances and perspectives of functionalized carbon dots in bacteria sensing.

Bacterial infectious diseases are severe threats to human health and increase substantial financial burdens. Nanomaterials have shown great potential in timely and accurate bacterial identification, detection, and monitoring to improve the cure rate and reduce mortality. Recently, carbon dots have been evidenced to be ideal candidates for bacterial identification and detection due to their superior physicochemical properties and biocompatibility. This review outlines the detailed recognition elements and recognition strategies with functionalized carbon dots (FCDs) for bacterial identification and detection. The advantages and limitations of different kinds of FCDs-based sensors will be critically discussed. Meanwhile, the ongoing challenges and perspectives of FCDs-based sensors for bacteria sensing are put forward.

The therapeutic alliance in couple therapy: Patterns by treatment and sex in a randomized controlled trial of emotionally focused therapy and treatment as usual.

The formation and development of the therapeutic alliance in couple therapy is a complex process and a key contributor to positive treatment outcomes. This study explored differences in trajectories of therapeutic alliance by sex and treatment condition among 24 couples randomized to receive Emotionally Focused Therapy or treatment as usual. The results identified a curvilinear growth pattern for alliance across both treatment groups. Female partners reported higher alliance than male partners after the first session across treatment groups, and female partners receiving Emotionally Focused Therapy reported higher initial alliance than female partners receiving treatment as usual. The rates of change for alliance did not differ by sex or treatment condition. The implications of the change pattern and differences in alliance formation by sex and treatment are discussed.

Sparse logistic regression revealed the associations between HBV PreS quasispecies and hepatocellular carcinoma.

BACKGROUND: Chronic infection with hepatitis B virus (HBV) has been proved highly associated with the development of hepatocellular carcinoma (HCC). AIMS: The purpose of the study is to investigate the association between HBV preS region quasispecies and HCC development, as well as to develop HCC diagnosis model using HBV preS region quasispecies. METHODS: A total of 104 chronic hepatitis B (CHB) patients and 117 HBV-related HCC patients were enrolled. HBV preS region was sequenced using next generation sequencing (NGS) and the nucleotide entropy was calculated for quasispecies evaluation. Sparse logistic regression (SLR) was used to predict HCC development and prediction performances were evaluated using receiver operating characteristic curves. RESULTS: Entropy of HBV preS1, preS2 regions and several nucleotide points showed significant divergence between CHB and HCC patients. Using SLR, the classification of HCC/CHB groups achieved a mean area under the receiver operating characteristic curve (AUC) of 0.883 in the training data and 0.795 in the test data. The prediction model was also validated by a completely independent dataset from Hong Kong. The 10 selected nucleotide positions showed significantly different entropy between CHB and HCC patients. The HBV quasispecies also classified three clinical parameters, including HBeAg, HBVDNA, and Alkaline phosphatase (ALP) with the AUC value greater than 0.6 in the test data. CONCLUSIONS: Using NGS and SLR, the association between HBV preS region nucleotide entropy and HCC development was validated in our study and this could promote the understanding of HCC progression mechanism.

Validation of the GALAD model for early diagnosis and monitoring of hepatocellular carcinoma in Chinese multicenter study.

BACKGROUND: GALAD is an algorithm model estimating the presence of hepatocellular carcinoma (HCC). However, the participants enrolled in the GALAD differ from those of Chinese subjects whose HCCs are mainly hepatitis B virus infection related. Therefore, the cross-sectional as well as longitudinal multicenter study was designed to assess the clinical performances of GALAD in the Chinese population. METHODS: A case-control study of 602 patients with HCC (34.10% within Barcelona Clinic Liver Cancer 0-A stage) and 923 subjects without HCC from five Chinese medical centres was conducted. Longitudinally the performances of GALAD identifying HCC were assessed using receiver operating characteristic curves analyses. Furthermore, the surveillance performance of GALAD for 204 HCC patients after radical surgery and for the early detection of HCC prospectively in an independent cohort of chronic hepatitis B were analysed, respectively. RESULTS: We found the GALAD identified early stage HCC at an area under the receiver operating characteristic curve (AUC) above 0.85 and outperformed significantly than AFP, PIVKAII, AFP-L3 and BALAD-2 respectively. Meanwhile the GALAD could stratify HCC into two distinct subgroups with high or low risks of overall survival and recurrence. The GALAD could detection HCC 24 (AUC: 0.848) or even 48 (AUC: 0.833) weeks before clinical diagnosis. CONCLUSIONS: Our study indicates that the GALAD exhibits outstanding performance in the early diagnosis, prognosis prediction as well as risk monitoring of HCC in our cross-sectional and longitudinal multicenter study of 1561 patients. GALAD should be implanted into clinical practice early so as to improve the clinical efficacy of individual biomarkers in HCC early monitoring and prognosis prediction.

Validation of the Eighth American Joint Committee on Cancer Anatomic and Prognostic Staging System for Breast Cancer.

BACKGROUND: The eighth edition of new staging systems for breast cancer incorporated four biological factors and the anatomic staging system. Validating analysis on Chinese patients has been limited. Our study performed analysis comparing the prognostic value of the staging system based on Chinese data. METHODS AND MATERIALS: All patients were classified according to the eighth edition and compared between anatomic and prognostic staging systems. The Kaplan-Meier test was used to calculate the overall survival (OS) and disease-free survival (DFS). We performed Harrell concordance index (C-index) analyses to quantify a models' predictive performance. Akaike information criterion (AIC) via Cox regression analysis was used to conduct bootstrap-based goodness-of-fit comparisons of the competing staging systems. RESULTS: A total of 1556 patients were enrolled in the cohort. The median follow-up time was 76 mo (range, 4-146 mo), the median age was 48 y old (range, 21-87 y). The ratio of movement between anatomic stage (AS) and prognostic stage (PS) was 50.9%. Of these, 691 (44.5%) AS patients were down staged and 100 (6.4%) patients were upstaged when reclassified based on PS. Significant differences between two stages were achieved for stage IIIC in 5-y OS rates and for IIIB in 5-y DFS rates (63.5% versus 50.0% and 58.0% versus44.0%). The value of the C-index for PS and AS were 0.711 and 0.687 (P = 0.04). The AIC reaches a value of 3452.9 for the PS and a value of 3476.4 for the AS. CONCLUSIONS: The PS might provide better accuracy than the AS in predicting the prognosis of Chinese female breast cancer patients. It also provides a strong basis for the utility of clinical biomarkers to evaluate the prognosis of patients.

Genetic studies discover novel coding and non-coding mutations in patients with Wilson's disease in China.

OBJECTIVES: Wilson disease (WD) is a rare autosomal recessive genetic disorder associated with various mutations in the ATP7B gene and leads to significant disability or death if untreated. Early diagnosis and proper therapy usually predict a good prognosis, especially in pre-symptomatic WD. Genetic testing provides an accurate and effective diagnostic method for the early diagnosis of WD. METHODS: We recruited 18 clinically diagnosed WD patients from 16 unrelated families and two independent individuals. The next-generation sequencing of the ATP7B gene was performed. The 293T cell lines were divided into wild-type (WT) ATP7B and mutated ATP7B groups. Cell proliferation was determined by Cell Counting Kit-8 (CCK-8) assay and apoptosis was detected by Annexin V/propidium iodide (PI) assays. RESULTS: Pedigree analysis showed that compound heterozygous variants (17/18, 94.44%) were present in the majority of WD patients. A total of 33 ATP7B gene variants were identified, including three variants with uncertain significance (VUS) [two splice mutations (c.51+2T>G, c.1543+40G>A) and one frameshift mutation (c.3532_3535del)]. The CCK-8 and apoptosis assays demonstrated that the VUS of ATP7B could significantly affect the transportation of copper. CONCLUSIONS: The study revealed genetic defects of 16 Chinese families and two independent individuals with WD, which enriched the mutation spectrum of the ATP7B gene worldwide and provided valuable information for studying the mutation types of ATP7B in the Chinese populations. Genetic testing in WD patients is necessary to shorten the time to initiate therapy, reduce damage to the liver and improve the prognosis.

Quantitative detections of TP53 gene mutations improve the diagnosis and prognostic prediction of biliary tract cancers using droplet digital PCR.

OBJECTIVE: Biliary tract cancer (BTC) is a rare malignancy and lack of effective diagnostic and prognostic marker. Here, we aimed to investigate the clinical implication of TP53 mutation detection in BTC using droplet digital PCR (ddPCR). METHODS: TP53 gene (loci p.R175H, p.R248Q, p.R248W, and p.R273H) mutation frequencies of 45 pairs of tumor tissues (TTs) and adjacent normal tissues (ANTTs) were analyzed, respectively, using ddPCR. Meanwhile, the same detections were conducted in plasma cell-free DNA (cfNDA) of 156 subjects including BTC, disease control (DC), and healthy controls (HC). The logistic regression algorithm was established to identify BTC. The correlations between mutations and clinicopathological features as well as the effects of TP53 mutation frequency on BTC prognosis were assessed. RESULTS: The higher mutation of p.R175H was found in TTs compared with ANTT (p = 0.006). The mutation at p.R273H in cfDNA was also higher in BTC when compared with DC and HC (p < 0.05). The logistic algorithms combining p.R273H mutation demonstrated the higher diagnostic efficacy trend than carbohydrate antigen 19-9 (CA19-9), carcinoembryonic antigen (CEA), and alpha-fetoprotein (AFP) in identifying BTC from DC (the area under the curves of the algorithm: 0.845, 95% CI:0.775-0.914). The median overall survival (OS) and progression-free survival (PFS) were significantly shorter when the BTC patients harboring the p.R273H mutation (OS: p = 0.032; PFS: p = 0.046). CONCLUSION: This study revealed for the first time that the quantitative TP53 mutations using the ddPCR might serve as a potential genetic biomarker for BTC diagnosis and prognosis assessment.

Comparative study on the main active components of Baoyuan decoction in plasma and urine of normal and heart failure rats.

The global morbidity and mortality of heart failure has been increasing in recent years. Traditional Chinese medicine (TCM) was increasingly used to treat cardiovascular diseases. Baoyuan decoction (BYD) was a famous classical prescription in China. Modern pharmacological studies showed that it had obvious therapeutic effects on cardiovascular diseases, but its pathological pharmacokinetic studies were unclear. In this research, the absorption of 16 bioactive components in plasma and the excretion of 9 representative components in urine of control rats and isoproterenol (ISO)-induced heart failure rats were studied using the large-volume direct-injection LC-MS method established by our research group. The results indicated that flavonoid constituents exhibited quicker absorption and elimination than saponin constituents after oral administration of BYD. The half-life period of some bioactive compounds in the model group was increased, which contributed to the longer therapeutic effect. The cumulative excretion rate of major flavonoid components of BYD decreased significantly in the ISO-induced heart failure rats.

Turning waste into wealth: nitrogen-doped carbon quantum dots derived from fruits wastes for sensing.

Hydrothermal treatment of m-phenylenediamine and grape seed powder has been adopted to synthesize nitrogen-doped carbon quantum dots (N-CQDs). The prepared N-CQDs possessed outstanding optical properties and high quantum yield. Based on the combined effect of static quenching effect and inner filter effect of permanganate (MnO4 - ) to N-CQDs and the redox reaction that occurred between MnO4 - and l-ascorbic acid (l-AA), an 'off-on' fluorescence strategy with N-CQDs has been proposed for the detection of MnO4 - and l-AA. The proposed fluorescent probe was fast, sensitive and selective to MnO4 - and l-AA under mild conditions. In addition, the satisfactory results of the proposed strategy for the detection of MnO4 - and l-AA in real samples indicated its practicability.

The Synergistic Mechanism of Total Saponins and Flavonoids in Notoginseng-Safflower against Myocardial Infarction Using a Comprehensive Metabolomics Strategy.

Notoginseng and safflower are commonly used traditional Chinese medicines for benefiting qi and activating blood circulation. A previous study by our group showed that the compatibility of the effective components of total saponins of notoginseng (NS) and total flavonoids of safflower (SF), named NS-SF, had a preventive effect on isoproterenol (ISO)-induced myocardial infarction (MI) in rats. However, the therapeutic effect on MI and the synergistic mechanism of NS-SF are still unclear. Therefore, integrated metabolomics, combined with immunohistochemistry and other pharmacological methods, was used to systematically research the therapeutic effect of NS-SF on MI rats and the synergistic mechanism of NS and SF. Compared to NS and SF, the results demonstrated that NS-SF exhibited a significantly better role in ameliorating myocardial damage, apoptosis, easing oxidative stress and anti-inflammation. NS-SF showed a more significant regulatory effect on metabolites involved in sphingolipid metabolism, glycine, serine, and threonine metabolism, primary bile acid biosynthesis, aminoacyl-tRNA biosynthesis, and tricarboxylic acid cycle, such as sphingosine, lysophosphatidylcholine (18:0), lysophosphatidylethanolamine (22:5/0:0), chenodeoxycholic acid, L-valine, glycine, and succinate, than NS or SF alone, indicating that NS and SF produced a synergistic effect on the treatment of MI. This study will provide a theoretical basis for the clinical development of NS-SF.

Semi-supervised Long-tail Endoscopic Image Classification.

Objective To explore the semi-supervised learning (SSL) algorithm for long-tail endoscopic image classification with limited annotations. Method We explored semi-supervised long-tail endoscopic image classification in HyperKvasir, the largest gastrointestinal public dataset with 23 diverse classes. Semi-supervised learning algorithm FixMatch was applied based on consistency regularization and pseudo-labeling. After splitting the training dataset and the test dataset at a ratio of 4:1, we sampled 20%, 50%, and 100% labeled training data to test the classification with limited annotations. Results The classification performance was evaluated by micro-average and macro-average evaluation metrics, with the Mathews correlation coefficient (MCC) as the overall evaluation. SSL algorithm improved the classification performance, with MCC increasing from 0.8761 to 0.8850, from 0.8983 to 0.8994, and from 0.9075 to 0.9095 with 20%, 50%, and 100% ratio of labeled training data, respectively. With a 20% ratio of labeled training data, SSL improved both the micro-average and macro-average classification performance; while for the ratio of 50% and 100%, SSL improved the micro-average performance but hurt macro-average performance. Through analyzing the confusion matrix and labeling bias in each class, we found that the pseudo-based SSL algorithm exacerbated the classifier's preference for the head class, resulting in improved performance in the head class and degenerated performance in the tail class. Conclusion SSL can improve the classification performance for semi-supervised long-tail endoscopic image classification, especially when the labeled data is extremely limited, which may benefit the building of assisted diagnosis systems for low-volume hospitals. However, the pseudo-labeling strategy may amplify the effect of class imbalance, which hurts the classification performance for the tail class.