Explore the vast range of titles available.

Background Predicting protein function and structure from sequence is one important challenge for computational biology. For 26 years, most state-of-the-art approaches combined machine learning and evolutionary information. However, for some applications retrieving related proteins is becoming too time-consuming. Additionally, evolutionary information is less powerful for small families, e.g. for proteins from the Dark Proteome . Both these problems are addressed by the new methodology introduced here. Results We introduced a novel way to represent protein sequences as continuous vectors ( embeddings ) by using the language model ELMo taken from natural language processing. By modeling protein sequences, ELMo effectively captured the biophysical properties of the language of life from unlabeled big data (UniRef50). We refer to these new embeddings as SeqVec ( Seq uence-to- Vec tor) and demonstrate their effectiveness by training simple neural networks for two different tasks. At the per-residue level, secondary structure (Q3 = 79% ± 1, Q8 = 68% ± 1) and regions with intrinsic disorder (MCC = 0.59 ± 0.03) were predicted significantly better than through one-hot encoding or through Word2vec-like approaches. At the per-protein level, subcellular localization was predicted in ten classes (Q10 = 68% ± 1) and membrane-bound were distinguished from water-soluble proteins (Q2 = 87% ± 1). Although SeqVec embeddings generated the best predictions from single sequences, no solution improved over the best existing method using evolutionary information. Nevertheless, our approach improved over some popular methods using evolutionary information and for some proteins even did beat the best. Thus, they prove to condense the underlying principles of protein sequences. Overall, the important novelty is speed: where the lightning-fast HHblits needed on average about two minutes to generate the evolutionary information for a target protein, SeqVec created embeddings on average in 0.03 s. As this speed-up is independent of the size of growing sequence databases, SeqVec provides a highly scalable approach for the analysis of big data in proteomics, i.e. microbiome or metaproteome analysis. Conclusion Transfer-learning succeeded to extract information from unlabeled sequence databases relevant for various protein prediction tasks. SeqVec modeled the language of life, namely the principles underlying protein sequences better than any features suggested by textbooks and prediction methods. The exception is evolutionary information, however, that information is not available on the level of a single sequence.

Large language models, in particular ChatGPT, have showcased remarkable language processing capabilities. Given the substantial workload of university medical staff, this study aims to assess the quality of multiple-choice questions (MCQs) produced by ChatGPT for use in graduate medical examinations, compared to questions written by university professoriate staffs based on standard medical textbooks. 50 MCQs were generated by ChatGPT with reference to two standard undergraduate medical textbooks (Harrison's, and Bailey & Love's). Another 50 MCQs were drafted by two university professoriate staff using the same medical textbooks. All 100 MCQ were individually numbered, randomized and sent to five independent international assessors for MCQ quality assessment using a standardized assessment score on five assessment domains, namely, appropriateness of the question, clarity and specificity, relevance, discriminative power of alternatives, and suitability for medical graduate examination. The total time required for ChatGPT to create the 50 questions was 20 minutes 25 seconds, while it took two human examiners a total of 211 minutes 33 seconds to draft the 50 questions. When a comparison of the mean score was made between the questions constructed by A.I. with those drafted by humans, only in the relevance domain that the A.I. was inferior to humans (A.I.: 7.56 +/- 0.94 vs human: 7.88 +/- 0.52; p = 0.04). There was no significant difference in question quality between questions drafted by A.I. versus humans, in the total assessment score as well as in other domains. Questions generated by A.I. yielded a wider range of scores, while those created by humans were consistent and within a narrower range. ChatGPT has the potential to generate comparable-quality MCQs for medical graduate examinations within a significantly shorter time.

A high-quality domain-oriented dataset is crucial for the domain-specific named entity recognition (NER) task. In this study, we introduce a novel education-oriented Chinese NER dataset (EduNER). To provide representative and diverse training data, we collect data from multiple sources, including textbooks, academic papers, and education-related web pages. The collected documents span ten years (2012–2021). A team of domain experts is invited to accomplish the education NER schema definition, and a group of trained annotators is hired to complete the annotation. A collaborative labeling platform is built for accelerating human annotation. The constructed EduNER dataset includes 16 entity types, 11k+ sentences, and 35,731 entities. We conduct a thorough statistical analysis of EduNER and summarize its distinctive characteristics by comparing it with eight open-domain or domain-specific NER datasets. Sixteen state-of-the-art models are further utilized for NER tasks validation. The experimental results can enlighten further exploration. To the best of our knowledge, EduNER is the first publicly available dataset for NER task in the education domain, which may promote the development of education-oriented NER models.

One of the aims of the present study was to reveal and compare the performance of ChatGPT versions (GPT-4o, GPT-4, and GPT-3.5), MathGPT, and Gemini in solving 390 mathematical problems in interactive mathematics e-textbooks across various dimensions. The other aim was to identify the affordances and constraints of ChatGPT through the instrumental approach. Both quantitative and qualitative approaches were used. The participants were 160 high school students and 80 mathematics teachers. Data were collected through the evaluation forms, view forms, and interviews with students and teachers based on ChatGPT’s solutions to the mathematical problems. The findings showed that the success rates of GPT-4o and GPT-4 were close to each other, with a slightly higher success rate of GPT-4o. This was followed by MathGPT and GPT-3.5. Gemini has the lowest success rate among the AI chatbots. Depending on the complexity of the mathematical problems, a statistically significant difference between the number of correct and incorrect solutions was found in all ChatGPT versions but not in MathGPT and Gemini. Furthermore, teachers and students referred to explanatory and detailed aspects of the solutions, learning without a teacher, getting solutions directly and quickly, and learning support as affordances of ChatGPT. On the other hand, the participants also acknowledged the constraints of ChatGPT while being aware of its affordances. Based on students’ and teachers’ expressions in the view forms and interviews, the mean scores they provided in the evaluation forms, and ChatGPT’s high performance in solving mathematical problems, it is suggested that ChatGPT could be a useful tool for students’ individual mathematics learning process.

Background and Objectives: Large language models (LLMs) are emerging as valuable tools in plastic surgery, potentially reducing surgeons’ cognitive loads and improving patients’ outcomes. This study aimed to assess and compare the current state of the two most common and readily available LLMs, Open AI’s ChatGPT-4 and Google’s Gemini Pro (1.0 Pro), in providing intraoperative decision support in plastic and reconstructive surgery procedures. Materials and Methods: We presented each LLM with 32 independent intraoperative scenarios spanning 5 procedures. We utilized a 5-point and a 3-point Likert scale for medical accuracy and relevance, respectively. We determined the readability of the responses using the Flesch–Kincaid Grade Level (FKGL) and Flesch Reading Ease (FRE) score. Additionally, we measured the models’ response time. We compared the performance using the Mann–Whitney U test and Student’s t-test. Results: ChatGPT-4 significantly outperformed Gemini in providing accurate (3.59 ± 0.84 vs. 3.13 ± 0.83, p-value = 0.022) and relevant (2.28 ± 0.77 vs. 1.88 ± 0.83, p-value = 0.032) responses. Alternatively, Gemini provided more concise and readable responses, with an average FKGL (12.80 ± 1.56) significantly lower than ChatGPT-4′s (15.00 ± 1.89) (p < 0.0001). However, there was no difference in the FRE scores (p = 0.174). Moreover, Gemini’s average response time was significantly faster (8.15 ± 1.42 s) than ChatGPT’-4′s (13.70 ± 2.87 s) (p < 0.0001). Conclusions: Although ChatGPT-4 provided more accurate and relevant responses, both models demonstrated potential as intraoperative tools. Nevertheless, their performance inconsistency across the different procedures underscores the need for further training and optimization to ensure their reliability as intraoperative decision-support tools.

Emotion detection from text is a relatively new sub-field of artificial intelligence closely related to Sentiment Analysis (SA). SA detects positive, neutral, or negative emotions in text. In contrast, emotion analysis detects and distinguishes certain types of emotions expressed in textbooks, such as disgust, fear, anger, happiness, surprise and sadness. Meanwhile, personality is a critical psychological concept that accounts for unique characteristics. Identifying and validating an individual’s personality efficiently and reliably is an admirable goal. This article aims to present a simultaneous review of Emotion and Personality detection from texts and elaborates upon approaches in developing text-based Emotion and Personality detection systems. The studies’ essential contributions, methodologies, datasets, conclusions drawn, strengths, and limitations are also explored. Additionally, this article discusses some of the field’s state-of-the-art ideas. In conclusion, the study delves into specific challenges and possible future research directions for detecting emotions and personalities from the text.

This study focuses on the translation and knowledge presentation of Chinese culture in art English textbooks. Due to the complex cultural context and highly specialized terminology in art English textbooks, traditional translation models struggle to accurately convey the deep semantic meaning and artistic value of Chinese culture. This paper proposes a translation enhancement method that integrates an improved Marian neural machine translation (Marian NMT) model with cultural adversarial reasoning networks (Cultural-Adversarial Reasoning Networks). The method employs transfer learning to incorporate Chinese cultural corpora for pre-training and combines a small amount of bilingual annotated data from art textbooks for fine-tuning. The model incorporates a cultural discriminator and generator adversarial mechanism to enhance the identification of culturally loaded words, art terminology, and context, thereby improving the cultural accuracy and educational suitability of the translation. Experiments were conducted on the “Chinese-English Parallel Corpus of Art English Textbooks,” covering themes such as painting, calligraphy, opera, and architecture. The results show that compared to the original Marian NMT, Transformer, and back-translation models, this method achieves significant improvements in BLEU, ROUGE, METEOR, and cultural knowledge integration accuracy (KIA), validating its effectiveness in translating Chinese cultural art English textbooks. The study concludes that this method can enhance the translation quality and teaching presentation effects of Chinese cultural elements in textbooks, providing technical support for the international dissemination of Chinese culture and textbook development.

Since chemistry, materials science, and crystallography deal with three-dimensional structures, they use mathematics such as geometry and symmetry. In recent years, the application of topology and mathematics to material design has yielded remarkable results. It can also be seen that differential geometry has been applied to various fields of chemistry for a relatively long time. There is also the possibility of using new mathematics, such as the crystal structure database, which represents big data, for computational chemistry (Hirshfeld surface analysis). On the other hand, group theory (space group and point group) is useful for crystal structures, including determining their electronic properties and the symmetries of molecules with relatively high symmetry. However, these strengths are not exhibited in the low-symmetry molecules that are actually handled. A new use of mathematics for chemical research is required that is suitable for the age when computational chemistry and artificial intelligence can be used.

[LANGUAGE= \"English\"] INTRODUCTION: Artificial intelligence (AI) has become the focus of recent studies, particularly due to its potential to reduce human labor and time loss. The most significant contribution of AI applications in the medical field is expected to be enhancing clinicians' efficiency, reducing costs, and improving public health. This study aims to assess the proficiency of ChatGPT-3.5, one of the most advanced AI applications available today, in its knowledge of current information based on the American Heart Association (AHA) 2020 guidelines.METHODS: An 80-question quiz in a question-and-answer format, covering the current AHA 2020 application steps, was prepared and administered to ChatGPT-3.5 in both English (ChatGPT-3.5 English) and Turkish (ChatGPT-3.5 Turkish). The questions were originally prepared in Turkish for emergency medicine specialists.RESULTS: We found a similar success rate of over 80% in all questions posed to ChatGPT-3.5 and two independent emergency medicine specialists with at least five years of experience who did not know each other. ChatGPT-3.5 achieved a 100% success rate in all questions related to the General Overview of the Current AHA Guidelines, Airway Management, and Ventilation chapters in English.DISCUSSION AND CONCLUSION: Our study indicates that ChatGPT-3.5 provides responses that are as accurate and up-to-date as those given by experienced emergency specialists regarding the AHA 2020 Advanced Cardiac Life Support Guidelines. With future updated versions of ChatGPT, instant access to accurate and current information based on textbooks and guidelines will be increasingly feasible.