Performance of ChatGPT on the Taiwan urology board examination: insights into current strengths and shortcomings.

PURPOSE: To compare ChatGPT-4 and ChatGPT-3.5's performance on Taiwan urology board examination (TUBE), focusing on answer accuracy, explanation consistency, and uncertainty management tactics to minimize score penalties from incorrect responses across 12 urology domains. METHODS: 450 multiple-choice questions from TUBE(2020-2022) were presented to two models. Three urologists assessed correctness and consistency of each response. Accuracy quantifies correct answers; consistency assesses logic and coherence in explanations out of total responses, alongside a penalty reduction experiment with prompt variations. Univariate logistic regression was applied for subgroup comparison. RESULTS: ChatGPT-4 showed strengths in urology, achieved an overall accuracy of 57.8%, with annual accuracies of 64.7% (2020), 58.0% (2021), and 50.7% (2022), significantly surpassing ChatGPT-3.5 (33.8%, OR = 2.68, 95% CI [2.05-3.52]). It could have passed the TUBE written exams if solely based on accuracy but failed in the final score due to penalties. ChatGPT-4 displayed a declining accuracy trend over time. Variability in accuracy across 12 urological domains was noted, with more frequently updated knowledge domains showing lower accuracy (53.2% vs. 62.2%, OR = 0.69, p = 0.05). A high consistency rate of 91.6% in explanations across all domains indicates reliable delivery of coherent and logical information. The simple prompt outperformed strategy-based prompts in accuracy (60% vs. 40%, p = 0.016), highlighting ChatGPT's limitations in its inability to accurately self-assess uncertainty and a tendency towards overconfidence, which may hinder medical decision-making. CONCLUSIONS: ChatGPT-4's high accuracy and consistent explanations in urology board examination demonstrate its potential in medical information processing. However, its limitations in self-assessment and overconfidence necessitate caution in its application, especially for inexperienced users. These insights call for ongoing advancements of urology-specific AI tools.

Using a Decision Tree Algorithm Predictive Model for Sperm Count Assessment and Risk Factors in Health Screening Population.

Purpose: Approximately 20% of couples face infertility challenges and struggle to conceive naturally. Despite advances in artificial reproduction, its success hinges on sperm quality. Our previous study used five machine learning (ML) algorithms, random forest, stochastic gradient boosting, least absolute shrinkage and selection operator regression, ridge regression, and extreme gradient boosting, to model health data from 1375 Taiwanese males and identified ten risk factors affecting sperm count. Methods: We employed the CART algorithm to generate decision trees using identified risk factors to predict healthy sperm counts. Four error metrics, SMAPE, RAE, RRSE, and RMSE, were used to evaluate the decision trees. We identified the top five decision trees based on their low errors and discussed in detail the tree with the least error. Results: The decision tree featuring the least error, comprising BMI, UA, ST, T-Cho/HDL-C ratio, and BUN, corroborated the negative impacts of metabolic syndrome, particularly high BMI, on sperm count, while emphasizing the link between good sleep and male fertility. Our study also sheds light on the potentially significant influence of high BUN on spermatogenesis. Two novel risk factors, T-Cho/HDL-C and UA, warrant further investigation. Conclusion: The ML algorithm established a predictive model for healthcare personnel to assess low sperm counts. Refinement of the model using additional data is crucial for improved precision. The risk factors identified offer avenues for future investigations.

Machine Learning Predictive Models for Evaluating Risk Factors Affecting Sperm Count: Predictions Based on Health Screening Indicators.

In many countries, especially developed nations, the fertility rate and birth rate have continually declined. Taiwan's fertility rate has paralleled this trend and reached its nadir in 2022. Therefore, the government uses many strategies to encourage more married couples to have children. However, couples marrying at an older age may have declining physical status, as well as hypertension and other metabolic syndrome symptoms, in addition to possibly being overweight, which have been the focus of the studies for their influences on male and female gamete quality. Many previous studies based on infertile people are not truly representative of the general population. This study proposed a framework using five machine learning (ML) predictive algorithms-random forest, stochastic gradient boosting, least absolute shrinkage and selection operator regression, ridge regression, and extreme gradient boosting-to identify the major risk factors affecting male sperm count based on a major health screening database in Taiwan. Unlike traditional multiple linear regression, ML algorithms do not need statistical assumptions and can capture non-linear relationships or complex interactions between dependent and independent variables to generate promising performance. We analyzed annual health screening data of 1375 males from 2010 to 2017, including data on health screening indicators, sourced from the MJ Group, a major health screening center in Taiwan. The symmetric mean absolute percentage error, relative absolute error, root relative squared error, and root mean squared error were used as performance evaluation metrics. Our results show that sleep time (ST), alpha-fetoprotein (AFP), body fat (BF), systolic blood pressure (SBP), and blood urea nitrogen (BUN) are the top five risk factors associated with sperm count. ST is a known risk factor influencing reproductive hormone balance, which can affect spermatogenesis and final sperm count. BF and SBP are risk factors associated with metabolic syndrome, another known risk factor of altered male reproductive hormone systems. However, AFP has not been the focus of previous studies on male fertility or semen quality. BUN, the index for kidney function, is also identified as a risk factor by our established ML model. Our results support previous findings that metabolic syndrome has negative impacts on sperm count and semen quality. Sleep duration also has an impact on sperm generation in the testes. AFP and BUN are two novel risk factors linked to sperm counts. These findings could help healthcare personnel and law makers create strategies for creating environments to increase the country's fertility rate. This study should also be of value to follow-up research.

TFEB- and TFE3-dependent autophagy activation supports cancer proliferation in the absence of centrosomes.

Centrosome amplification is a phenomenon frequently observed in human cancers, so centrosome depletion has been proposed as a therapeutic strategy. However, despite being afflicted with a lack of centrosomes, many cancer cells can still proliferate, implying there are impediments to adopting centrosome depletion as a treatment strategy. Here, we show that TFEB- and TFE3-dependent autophagy activation contributes to acentrosomal cancer proliferation. Our biochemical analyses uncover that both TFEB and TFE3 are novel PLK4 (polo like kinase 4) substrates. Centrosome depletion inactivates PLK4, resulting in TFEB and TFE3 dephosphorylation and subsequent promotion of TFEB and TFE3 nuclear translocation and transcriptional activation of autophagy- and lysosome-related genes. A combination of centrosome depletion and inhibition of the TFEB-TFE3 autophagy-lysosome pathway induced strongly anti-proliferative effects in cancer cells. Thus, our findings point to a new strategy for combating cancer.Abbreviations: AdCre: adenoviral Cre recombinase; AdLuc: adenoviral luciferase; ATG5: autophagy related 5; CQ: chloroquine; DAPI: 4',6-diamidino-2-phenylindole; DKO: double knockout; GFP: green fluorescent protein; KO: knockout; LAMP1: lysosomal associated membrane protein 1; LAMP2: lysosomal associated membrane protein 2; LTR: LysoTracker Red; MAP1LC3B/LC3B: microtubule associated protein 1 light chain 3 beta; MITF: melanocyte inducing transcription factor; PLK4: polo like kinase 4; RFP: red fluorescent protein; SASS6: SAS-6 centriolar assembly protein; STIL: STIL centriolar assembly protein; TFEB: transcription factor EB; TFEBΔNLS: TFEB lacking a nuclear localization signal; TFE3: transcription factor binding to IGHM enhancer 3; TP53/p53: tumor protein p53.

GnT1IP-L specifically inhibits MGAT1 in the Golgi via its luminal domain.

Mouse GnT1IP-L, and membrane-bound GnT1IP-S (MGAT4D) expressed in cultured cells inhibit MGAT1, the N-acetylglucosaminyltransferase that initiates the synthesis of hybrid and complex N-glycans. However, it is not known where in the secretory pathway GnT1IP-L inhibits MGAT1, nor whether GnT1IP-L inhibits other N-glycan branching N-acetylglucosaminyltransferases of the medial Golgi. We show here that the luminal domain of GnT1IP-L contains its inhibitory activity. Retention of GnT1IP-L in the endoplasmic reticulum (ER) via the N-terminal region of human invariant chain p33, with or without C-terminal KDEL, markedly reduced inhibitory activity. Dynamic fluorescent resonance energy transfer (FRET) and bimolecular fluorescence complementation (BiFC) assays revealed homomeric interactions for GnT1IP-L in the ER, and heteromeric interactions with MGAT1 in the Golgi. GnT1IP-L did not generate a FRET signal with MGAT2, MGAT3, MGAT4B or MGAT5 medial Golgi GlcNAc-tranferases. GnT1IP/Mgat4d transcripts are expressed predominantly in spermatocytes and spermatids in mouse, and are reduced in men with impaired spermatogenesis.

A testis-specific regulator of complex and hybrid N-glycan synthesis.

Database analyses identified 4933434I20Rik as a glycosyltransferase-like gene expressed mainly in testicular germ cells and regulated during spermatogenesis. Expression of a membrane-bound form of the protein resulted in a marked and specific reduction in N-acetylglucosaminyltransferase I (GlcNAcT-I) activity and complex and hybrid N-glycan synthesis. Thus, the novel activity was termed GlcNAcT-I inhibitory protein (GnT1IP). Membrane-bound GnT1IP localizes to the ER, the ER-Golgi intermediate compartment (ERGIC), and the cis-Golgi. Coexpression of membrane-anchored GnT1IP with GlcNAcT-I causes association of the two proteins, inactivation of GlcNAcT-I, and mislocalization of GlcNAcT-I from the medial-Golgi to earlier compartments. Therefore, GnT1IP is a regulator of GlcNAcT-I and complex and hybrid N-glycan production. Importantly, the formation of high mannose N-glycans resulting from inhibition of GlcNAcT-I by GnT1IP markedly increases the adhesion of CHO cells to TM4 Sertoli cells. Testicular germ cells might use GnT1IP to induce the expression of high mannose N-glycans on glycoproteins, thereby facilitating Sertoli-germ cell attachment at a particular stage of spermatogenesis.

Glycomics profiling of Chinese hamster ovary cell glycosylation mutants reveals N-glycans of a novel size and complexity.

Identifying biological roles for mammalian glycans and the pathways by which they are synthesized has been greatly facilitated by investigations of glycosylation mutants of cultured cell lines and model organisms. Chinese hamster ovary (CHO) glycosylation mutants isolated on the basis of their lectin resistance have been particularly useful for glycosylation engineering of recombinant glycoproteins. To further enhance the application of these mutants, and to obtain insights into the effects of altering one specific glycosyltransferase or glycosylation activity on the overall expression of cellular glycans, an analysis of the N-glycans and major O-glycans of a panel of CHO mutants was performed using glycomic analyses anchored by matrix-assisted laser desorption ionization-time of flight/time of flight mass spectrometry. We report here the complement of the major N-glycans and O-glycans present in nine distinct CHO glycosylation mutants. Parent CHO cells grown in monolayer versus suspension culture had similar profiles of N- and O-GalNAc glycans, although the profiles of glycosylation mutants Lec1, Lec2, Lec3.2.8.1, Lec4, LEC10, LEC11, LEC12, Lec13, and LEC30 were consistent with available genetic and biochemical data. However, the complexity of the range of N-glycans observed was unexpected. Several of the complex N-glycan profiles contained structures of m/z approximately 13,000 representing complex N-glycans with a total of 26 N-acetyllactosamine (Gal beta1-4GlcNAc)(n) units. Importantly, the LEC11, LEC12, and LEC30 CHO mutants exhibited unique complements of fucosylated complex N-glycans terminating in Lewis(x) and sialyl-Lewis(x) determinants. This analysis reveals the larger-than-expected complexity of N-glycans in CHO cell mutants that may be used in a broad variety of functional glycomics studies and for making recombinant glycoproteins.

N-Glycan structures of squid rhodopsin.

To determine the glycoforms of squid rhodopsin, N-glycans were released by glycoamidase A digestion, reductively aminated with 2-aminopyridine, and then subjected to 2D HPLC analysis [Takahashi, N., Nakagawa, H., Fujikawa, K., Kawamura, Y. & Tomiya, N. (1995) Anal. Biochem.226, 139-146]. The major glycans of squid rhodopsin were shown to possess the alpha1-3 and alpha1-6 difucosylated innermost GlcNAc residue found in glycoproteins produced by insects and helminths. By combined use of 2D HPLC, electrospray ionization-mass spectrometry and permethylation and gas chromatography-electron ionization mass spectrometry analyses, it was revealed that most (85%) of the N-glycans exhibit the novel structure Manalpha1-6(Manalpha1-3)Manbeta1-4GlcNAcbeta1-4(Galbeta1-4Fucalpha1-6)(Fucalpha1-3)GlcNAc.