• 【机器学习和生物医学领域的单词歧义消除:设计和评估问题。】 复制标题 收藏 收藏
    DOI:10.1186/1471-2105-7-334 复制DOI
    作者列表:Xu H,Markatou M,Dimova R,Liu H,Friedman C
    BACKGROUND & AIMS: BACKGROUND:Word sense disambiguation (WSD) is critical in the biomedical domain for improving the precision of natural language processing (NLP), text mining, and information retrieval systems because ambiguous words negatively impact accurate access to literature containing biomolecular entities, such as genes, proteins, cells, diseases, and other important entities. Automated techniques have been developed that address the WSD problem for a number of text processing situations, but the problem is still a challenging one. Supervised WSD machine learning (ML) methods have been applied in the biomedical domain and have shown promising results, but the results typically incorporate a number of confounding factors, and it is problematic to truly understand the effectiveness and generalizability of the methods because these factors interact with each other and affect the final results. Thus, there is a need to explicitly address the factors and to systematically quantify their effects on performance. RESULTS:Experiments were designed to measure the effect of "sample size" (i.e. size of the datasets), "sense distribution" (i.e. the distribution of the different meanings of the ambiguous word) and "degree of difficulty" (i.e. the measure of the distances between the meanings of the senses of an ambiguous word) on the performance of WSD classifiers. Support Vector Machine (SVM) classifiers were applied to an automatically generated data set containing four ambiguous biomedical abbreviations: BPD, BSA, PCA, and RSV, which were chosen because of varying degrees of differences in their respective senses. Results showed that: 1) increasing the sample size generally reduced the error rate, but this was limited mainly to well-separated senses (i.e. cases where the distances between the senses were large); in difficult cases an unusually large increase in sample size was needed to increase performance slightly, which was impractical, 2) the sense distribution did not have an effect on performance when the senses were separable, 3) when there was a majority sense of over 90%, the WSD classifier was not better than use of the simple majority sense, 4) error rates were proportional to the similarity of senses, and 5) there was no statistical difference between results when using a 5-fold or 10-fold cross-validation method. Other issues that impact performance are also enumerated. CONCLUSION:Several different independent aspects affect performance when using ML techniques for WSD. We found that combining them into one single result obscures understanding of the underlying methods. Although we studied only four abbreviations, we utilized a well-established statistical method that guarantees the results are likely to be generalizable for abbreviations with similar characteristics. The results of our experiments show that in order to understand the performance of these ML methods it is critical that papers report on the baseline performance, the distribution and sample size of the senses in the datasets, and the standard deviation or confidence intervals. In addition, papers should also characterize the difficulty of the WSD task, the WSD situations addressed and not addressed, as well as the ML methods and features used. This should lead to an improved understanding of the generalizablility and the limitations of the methodology.
    背景与目标: 背景:字词歧义消除(WSD)在生物医学领域对于提高自然语言处理(NLP),文本挖掘和信息检索系统的精度至关重要,因为模棱两可的单词会对准确访问包含生物分子实体(例如基因)的文献产生负面影响蛋白质,细胞,疾病和其他重要实体。已经开发出自动技术来解决许多文本处理情况下的WSD问题,但是该问题仍然是一个具有挑战性的问题。有监督的WSD机器学习(ML)方法已应用于生物医学领域,并显示出令人鼓舞的结果,但是结果通常包含许多混杂因素,并且由于这些因素相互作用,真正了解这些方法的有效性和可推广性是有问题的彼此影响最终结果。因此,需要明确解决这些因素并系统地量化其对性能的影响。
    结果:设计了实验来测量“样本量”(即数据集的大小),“感官分布”(即歧义词的不同含义的分布)和“难易程度”(即歧义词的含义之间的距离)对WSD分类器的性能。支持向量机(SVM)分类器应用于自动生成的数据集,该数据集包含四个歧义生物医学缩写:BPD,BSA,PCA和RSV,这是由于它们各自含义上的差异程度不同而选择的。结果表明:1)增加样本大小通常会降低错误率,但这主要限于良好分离的感官(即,感官之间的距离较大的情况);在困难的情况下,需要极大地增加样本数量以略微提高性能,这是不切实际的; 2)当感官可分离时,感官分布对性能没有影响; 3)当多数感官超过90时%,WSD分类器并不比使用简单多数感官更好; 4)错误率与各种感官的相似性成正比; 5)当使用5倍或10倍交叉比对时,结果之间没有统计学差异验证方法。还列举了影响性能的其他问题。
    结论:将ML技术用于WSD时,有几个不同的独立方面会影响性能。我们发现将它们组合成一个单一的结果会模糊对基本方法的理解。尽管我们仅研究了四个缩写,但我们使用了一种完善的统计方法,该方法可以保证结果对于具有相似特征的缩写很可能是可推广的。我们的实验结果表明,为了了解这些ML方法的性能,至关重要的是,论文要报告基线性能,数据集中感官的分布和样本大小以及标准偏差或置信区间。此外,论文还应描述WSD任务的难度,WSD解决和未解决的WSD情况以及所使用的ML方法和功能。这应该导致人们对通用性和方法的局限性有了更好的了解。
  • 【广泛耐药结核病:我们是在学习历史还是在重复历史?】 复制标题 收藏 收藏
    DOI:10.1086/519292 复制DOI
    作者列表:Dukes Hamilton C,Sterling TR,Blumberg HM,Leonard M,McAuley J,Schlossberg D,Stout J,Huitt G
    BACKGROUND & AIMS: :Tuberculosis (TB) is an enormous global public health problem. Cases of extensively drug-resistant TB (XDR-TB) are being reported in increasing numbers across the globe. A large outbreak of XDR-TB associated with rapid and nearly universal mortality has been reported among patients with human immunodeficiency virus infection or acquired immunodeficiency disease in South Africa who have been receiving standard TB therapy and antiretrovirals. Epidemiologic features of this outbreak make it highly suspicious for health care-associated transmission. We urge the Infectious Diseases Society of America and its members to increase involvement in ongoing international TB prevention and treatment efforts and to develop a registry of experts in infection control and laboratory and disease management. We urge advocacy for increased funding for domestic and global TB control programs, including expanded access to sputum culture and drug susceptibility testing, as well as funding for TB clinical trials and research capacity. We believe that substandard TB diagnostic tests are not acceptable for TB control in resource-poor countries. We urge the development of shorter, less toxic TB treatment and prevention regimens. Funding of TB control and research should be reassessed to prevent budget cuts at a time when the disease is killing as many as 2 million people a year.
    背景与目标: 结核病(TB)是一个巨大的全球公共卫生问题。在全球范围内,越来越多的人报告了广泛耐药的结核病(XDR-TB)病例。在南非,已经接受标准结核病治疗和抗逆转录病毒药物治疗的人类免疫缺陷病毒感染或获得性免疫缺陷疾病患者中,广泛的XDR-TB暴发已被报道。该暴发的流行病学特征使其高度怀疑与卫生保健有关的传播。我们敦促美国传染病学会及其成员增加对正在进行的国际结核病预防和治疗工作的参与,并发展感染控制以及实验室和疾病管理方面的专家名录。我们敦促倡导为国内和全球结核病控制计划增加资金,包括扩大对痰培养和药物敏感性测试的访问,以及为结核病临床试验和研究能力提供资金。我们认为,在资源匮乏的国家,不合格的结核病诊断检测对于控制结核病是不可接受的。我们敦促开发更短,毒性更小的结核病治疗和预防方案。在结核病每年导致多达200万人死亡的时候,应该重新评估结核病控制和研究的资金,以防止预算削减。
  • 【收集经验丰富的专业知识来支持糖尿病患者的安全驾驶:一项由同伴在一项调查中评估的定性研究。】 复制标题 收藏 收藏
    DOI:10.1007/BF03262497 复制DOI
    作者列表:Burda MH,van der Horst F,van den Akker M,Stork AD,Mesters I,Bours S,Ploeg M,Winkens B,Knottnerus JA
    BACKGROUND & AIMS: BACKGROUND:Hypoglycemia is a frequent phenomenon in people being treated for diabetes mellitus, which can acutely disrupt driving performance. For the benefit of personal and public traffic safety, we decided to identify successful diabetes-related (SDR) behaviors to support safe driving for people with diabetes, from the perspective of experiential experts with diabetes mellitus. Experiential experts are people who can manage their own illness and conditions by developing expertise relevant to maintaining health and countering illness, and who are able to use this expertise to the benefit of peers. OBJECTIVE:The aim of our study was to objectify and systematize experiential expertise in terms of SDR behaviors, based on reports by experiential experts, to support safe driving for people with type 1 and type 2 diabetes mellitus. The emphasis was on preventing hypoglycemia as a short-term complication during driving. METHODS:We performed a mixed-methods study involving (i) semi-structured in-depth interviews with 33 experiential experts with diabetes mellitus from the Dutch Diabetes Association (DVN; Diabetesvereniging Nederland), in order to identify SDR behaviors regarding safe driving, and (ii) a validation study by means of a survey among a panel of 98 experiential experts (peers) from the DVN, to determine the extent to which they agreed with the communicability, importance, and feasibility of these behaviors for drivers with diabetes mellitus. RESULTS:We identified a comprehensive set of 11 SDR behaviors, differentiated into seven general and four specific behaviors, to support safe driving. The general behaviors concern the following topics: (i) acquiring knowledge and information; (ii) acquiring and using self-measuring of blood glucose (SMBG) equipment; (iii) knowing one's physical response pattern; (iv) obtaining knowledge about the medication used; (v) preventing long-term eye complications; (vi) influencing factors that can affect blood glucose; and (vii) renewal procedure for driving license. The four specific behaviors refer to the following topics: (i) measures to be taken before driving; (ii) responding effectively to hypoglycemia while driving; (iii) informing and instructing passengers; and (iv) preventing hypoglycemia in drivers with type 2 diabetes mellitus not using SMBG equipment. Key factors for safe driving proved to be the ability of drivers to anticipate and respond effectively to hypoglycemia while driving and to inform and instruct fellow passengers. Participants of the validation survey agreed to a considerable degree with the communicability, importance, and feasibility of these behaviors to support safe driving for people with diabetes mellitus. CONCLUSIONS:This study resulted in the identification and description of SDR behaviors to support safe driving. It proved possible to operationalize experiential expertise in terms of such behaviors. The next step is to have these behaviors validated by professional care providers in the field of diabetes, followed by translation into recommendations in self-management programs.
    背景与目标: 背景:低血糖症在接受糖尿病治疗的人们中很常见,会严重破坏驾驶性能。为了个人和公共交通安全,我们决定从经验丰富的糖尿病专家的角度确定成功的糖尿病相关(SDR)行为,以支持糖尿病人安全驾驶。经验丰富的专家可以通过发展与维护健康和对抗疾病有关的专业知识来管理自己的疾病和状况,并能够利用这些专业知识来为同行带来好处。
    目的:我们的研究目的是根据经验专家的报告,对SDR行为进行客观化和系统化的专门知识,以支持1型和2型糖尿病患者的安全驾驶。重点是预防驾驶过程中的短期并发症即低血糖症。
    方法:我们进行了一项混合方法研究,涉及(i)对来自荷兰糖尿病协会(DVN; Diabetesvereniging Nederland)的33位糖尿病专家进行的半结构化深度访谈,以确定与安全驾驶有关的SDR行为,以及(ii)通过对来自DVN的98位经验丰富的专家(同行)进行的调查来进行验证研究,以确定他们在多大程度上同意这些行为对于糖尿病驾驶员的可传播性,重要性和可行性。
    结果:我们确定了11种SDR行为的综合集合,分为7种一般行为和4种特定行为,以支持安全驾驶。一般行为涉及以下主题:(i)获取知识和信息; (ii)获取和使用自我测量的血糖仪(SMBG); (iii)了解一个人的身体反应方式; (iv)获得有关所用药物的知识; (v)预防眼部长期并发症; (vi)可能影响血糖的影响因素; (vii)驾驶执照的续签程序。四种具体行为涉及以下主题:(i)驾驶前应采取的措施; (ii)驾驶时有效应对低血糖; (iii)通知和指示乘客; (iv)预防不使用SMBG设备的2型糖尿病驾驶员的低血糖症。事实证明,安全驾驶的关键因素是驾驶员在驾驶时能够预见和有效应对低血糖以及告知和指导同伴的能力。验证调查的参与者在相当程度上同意这些行为的可传播性,重要性和可行性,以支持糖尿病人的安全驾驶。
    结论:本研究确定并描述了支持安全驾驶的SDR行为。事实证明,可以根据这种行为来操作经验专业知识。下一步是使这些行为得到糖尿病领域专业护理人员的验证,然后转化为自我管理计划中的建议。
  • 【向饥饿的学习者分发菜单:通过模拟学习可以成为学习的模拟吗?】 复制标题 收藏 收藏
    DOI:10.1080/01421590601042335 复制DOI
    作者列表:Bligh J,Bleakley A
    BACKGROUND & AIMS: :Simulation offers an important context for clinical education, providing a structured, safe and supportive environment bridging the classroom and the clinic. Two trends in the simulation community appear to be developing uncritically and without adequate evaluation. First, there is a fascination with seductive high-fidelity simulation realized through sophisticated technology. Second, simulation has increasingly appropriated learning in the psychological domain, such as communication skills, under the rationale of 'integration'. Developments in simulation activities have largely been made in a theoretical vacuum and where theory is invoked it is learning theory rather than theory of simulation. This paper introduces theories of simulation from cultural studies as a critical balance to the claims of the simulation community. Work-based and simulation-based learning could engage in a new dialogue for an effective clinical education.
    背景与目标: :模拟为临床教育提供了重要的环境,为架设教室和诊所提供了结构化,安全和支持性的环境。在模拟社区中,有两种趋势似乎正在不加批判地发展,并且没有进行充分的评估。首先,对通过复杂技术实现的诱人的高保真度模拟着迷。其次,在“整合”的基本原理下,模拟越来越适合在心理学领域中进行学习,例如交流技巧。模拟活动的发展很大程度上是在理论真空中进行的,在其中调用理论的是学习理论而不是模拟理论。本文介绍了文化研究中的模拟理论,作为对模拟社区主张的重要平衡。基于工作和基于模拟的学习可以参与新的对话,以进行有效的临床教育。
  • 【P物质及其在控制学习,焦虑和功能恢复的神经机制中的作用。】 复制标题 收藏 收藏
    DOI:10.1054/npep.2000.0824 复制DOI
    作者列表:Hasenöhrl RU,Souza-Silva MA,Nikolaus S,Tomaz C,Brandao ML,Schwarting RK,Huston JP
    BACKGROUND & AIMS: :The neurokinin Substance P (SP) is widely distributed in the central nervous system and has been extensively studied in various functional aspects. This review focuses on the behavioral relevance of SP. Here we show that SP can have memory-promoting, reinforcing and anxiolytic-like effects when administered systemically or into the nucleus basalis of the ventral pallidum. These effects seem to be mediated via the SP-preferring NK(1)receptor and differentially related to N- versus C-terminal fragments of the undecapeptide. Secondly, SP injection into the ventral pallidum can lead to increases of acetylcholine in frontal cortex and dopamine in nucleus accumbens, suggesting that the hypermnestic, positively reinforcing and anxiolytic effects observed upon basal forebrain injection of SP are mediated by activation of the nucleus accumbens-ventral pallidum circuitry. Furthermore, SP and certain SP-fragments may not only be considered to have beneficial behavioral effects in normal animals, but can also prevent lesion-induced functional deficits and improve the speed of recovery. This indicates that SP agonists might also have a neuroprotective capacity in parallel with recovery-promoting actions.
    背景与目标: :神经激肽P(SP)广泛分布于中枢神经系统,并已在各个功能方面进行了广泛研究。这篇综述着重于SP的行为相关性。在这里我们表明,当全身或腹侧苍白球的核基底层给药时,SP可以具有促进记忆,增强和抗焦虑的作用。这些作用似乎是通过SP优先的NK(1)受体介导的,并且与十一肽的N端和C端片段差异相关。其次,向腹侧苍白球注射SP可导致额叶皮层中的乙酰胆碱和伏隔核中的多巴胺增加,表明基础前脑注射SP时观察到的高记忆力,正增强和抗焦虑作用是由伏隔核-腹侧神经的激活介导的。苍白的电路。此外,SP和某些SP片段不仅可以被认为在正常动物中具有有益的行为效果,而且还可以防止病变引起的功能缺陷并提高恢复速度。这表明SP激动剂可能还具有促进恢复作用的神经保护能力。
  • 【经颅多天的直流电刺激改善了新词汇的学习和维持。】 复制标题 收藏 收藏
    DOI:10.1016/j.cortex.2013.07.013 复制DOI
    作者列表:Meinzer M,Jähnigen S,Copland DA,Darkow R,Grittner U,Avirame K,Rodriguez AD,Lindenberg R,Flöel A
    BACKGROUND & AIMS: INTRODUCTION:Recently, growing interest emerged in the enhancement of human potential by means of non-invasive brain stimulation. In particular, anodal transcranial direct current stimulation (atDCS) has been shown to exert beneficial effects on motor and higher cognitive functions. However, the majority of transcranial direct current stimulation (tDCS) studies have assessed effects of single stimulation sessions that are mediated by transient neural modulation. Studies assessing the impact of multiple stimulation sessions on learning that may induce long-lasting behavioural and neural changes are scarce and have not yet been accomplished in the language domain in healthy individuals. METHOD:The present study probed the potential of atDCS to enhance language learning over multiple days by employing an explicit word learning paradigm. Forty healthy young participants were randomized to learning with either simultaneous atDCS or sham stimulation (N = 20/group; comparable regarding demographic variables and neurocognitive status). All participants acquired a novel vocabulary (familiar and novel object picture - non-word pairs) over five consecutive days. Two memory tasks (free recall; forced choice recognition tasks) were administered immediately after each training session. A one week follow-up tested the maintenance of learning success. RESULTS:Linear mixed effects model analysis revealed superior learning during atDCS compared to sham stimulation for both familiar and novel objects. atDCS yielded a steeper learning curve and significantly more pronounced learning at the end of the training during the recall task. During the recognition task, the atDCS group reached ceiling levels earlier and overall learning success was greater. For both tasks, beneficial atDCS effects were maintained during the follow-up assessment. CONCLUSIONS:The present study provides direct evidence that atDCS administered during multiple learning sessions facilitates language learning and that effects are maintained over time. This study contributes important novel information about the extent of stimulation effects in the healthy brain, thereby highlighting the potential of atDCS to enhance language recovery after stroke.
    背景与目标: 简介:最近,人们越来越关注通过无创性脑刺激来增强人类潜能。特别是,阳极经颅直流电刺激(atDCS)已显示出对运动和更高的认知功能产生有益作用。但是,大多数经颅直流电刺激(tDCS)研究已经评估了由短暂神经调节介导的单个刺激阶段的效果。评估多种刺激对学习的影响的研究可能会导致持久的行为和神经变化,但目前尚缺乏健康人在语言领域进行的研究,这些研究尚未完成。
    方法:本研究探讨了atDCS通过采用显式单词学习范例在多天内增强语言学习的潜力。四十名健康的年轻参与者被随机分配到同时进行DCS或假刺激的学习中(N = 20 /组;在人口统计学变量和神经认知状态方面具有可比性)。所有参与者都连续五天获得了新颖的词汇表(熟悉的事物和新颖的事物图片-非单词对)。每次训练后,立即执行两个记忆任务(自由回忆;强制选择识别任务)。为期一周的随访测试了学习成功的维持情况。
    结果:线性混合效应模型分析显示,与熟悉刺激和新颖物体的假刺激相比,atDCS期间的学习效果更好。在召回任务期间的培训结束时,atDCS产生了更陡峭的学习曲线,并且学习效果明显提高。在识别任务期间,atDCS组更早达到最高级别,并且整体学习成功更大。对于这两项任务,在后续评估期间都保持了atDCS的有益效果。
    结论:本研究提供了直接的证据,表明在多次学习中使用atDCS可以促进语言学习,并且效果会随着时间的推移而保持。这项研究提供了有关健康大脑中刺激作用程度的重要新信息,从而突出了atDCS增强中风后语言恢复的潜力。
  • 【培训麻醉学住院医师为清醒的开颅手术提供麻醉:学习曲线并估计所需病例数。】 复制标题 收藏 收藏
    DOI:10.1016/j.jclinane.2013.01.012 复制DOI
    作者列表:Bilotta F,Titi L,Lanni F,Stazi E,Rosa G
    BACKGROUND & AIMS: STUDY OBJECTIVE:To measure the learning curves of residents in anesthesiology in providing anesthesia for awake craniotomy, and to estimate the case load needed to achieve a "good-excellent" level of competence. DESIGN:Prospective study. SETTING:Operating room of a university hospital. SUBJECTS:7 volunteer residents in anesthesiology. MEASUREMENTS:Residents underwent a dedicated training program of clinical characteristics of anesthesia for awake craniotomy. The program was divided into three tasks: local anesthesia, sedation-analgesia, and intraoperative hemodynamic management. The learning curve for each resident for each task was recorded over 10 procedures. Quantitative assessment of the individual's ability was based on the resident's self-assessment score and the attending anesthesiologist's judgment, and rated by modified 12 mm Likert scale, reported ability score visual analog scale (VAS). This ability VAS score ranged from 1 to 12 (ie, very poor, mild, moderate, sufficient, good, excellent). The number of requests for advice also was recorded (ie, resident requests for practical help and theoretical notions to accomplish the procedures). MAIN RESULTS:Each task had a specific learning rate; the number of procedures necessary to achieve "good-excellent" ability with confidence, as determined by the recorded results, were 10 procedures for local anesthesia, 15 to 25 procedures for sedation-analgesia, and 20 to 30 procedures for intraoperative hemodynamic management. CONCLUSIONS:Awake craniotomy is an approach used increasingly in neuroanesthesia. A dedicated training program based on learning specific tasks and building confidence with essential features provides "good-excellent" ability.
    背景与目标: 目的:测量麻醉学中为清醒开颅手术提供麻醉的居民的学习曲线,并评估达到“卓越”水平所需的病案负荷。
    设计:前瞻性研究。
    地点:大学医院的手术室。
    受试者:7名麻醉学志愿者。
    测量:居民接受了清醒开颅手术麻醉临床特征的专门培训计划。该程序分为三个任务:局部麻醉,镇静镇痛和术中血流动力学管理。在10个步骤中记录了每个居民每个任务的学习曲线。对个人能力的定量评估基于居民的自我评估得分和主治麻醉师的判断,并通过改良的12毫米李克特量表,报告的能力得分视觉模拟量表(VAS)进行评估。此功能的VAS分数介于1到12之间(即非常差,中等,中等,足够,良好,非常好)。还记录了咨询请求的数量(即居民要求提供实际帮助和完成程序的理论观念)。
    主要结果:每项任务都有特定的学习率。根据记录的结果确定,要有信心达到“优秀”能力所必需的操作数为:局部麻醉10例,镇静镇痛15到25例,术中血流动力学管理20到30例。
    结论:清醒颅骨切开术是一种越来越多地用于神经麻醉的方法。基于学习特定任务并建立具有基本特征的自信心的专门培训计划可提供“出色”的能力。
  • 【基于深度学习方法的赖氨酸琥珀酰化位点的表征和鉴定。】 复制标题 收藏 收藏
    DOI:10.1038/s41598-019-52552-4 复制DOI
    作者列表:Huang KY,Hsu JB,Lee TY
    BACKGROUND & AIMS: :Succinylation is a type of protein post-translational modification (PTM), which can play important roles in a variety of cellular processes. Due to an increasing number of site-specific succinylated peptides obtained from high-throughput mass spectrometry (MS), various tools have been developed for computationally identifying succinylated sites on proteins. However, most of these tools predict succinylation sites based on traditional machine learning methods. Hence, this work aimed to carry out the succinylation site prediction based on a deep learning model. The abundance of MS-verified succinylated peptides enabled the investigation of substrate site specificity of succinylation sites through sequence-based attributes, such as position-specific amino acid composition, the composition of k-spaced amino acid pairs (CKSAAP), and position-specific scoring matrix (PSSM). Additionally, the maximal dependence decomposition (MDD) was adopted to detect the substrate signatures of lysine succinylation sites by dividing all succinylated sequences into several groups with conserved substrate motifs. According to the results of ten-fold cross-validation, the deep learning model trained using PSSM and informative CKSAAP attributes can reach the best predictive performance and also perform better than traditional machine-learning methods. Moreover, an independent testing dataset that truly did not exist in the training dataset was used to compare the proposed method with six existing prediction tools. The testing dataset comprised of 218 positive and 2621 negative instances, and the proposed model could yield a promising performance with 84.40% sensitivity, 86.99% specificity, 86.79% accuracy, and an MCC value of 0.489. Finally, the proposed method has been implemented as a web-based prediction tool (CNN-SuccSite), which is now freely accessible at http://csb.cse.yzu.edu.tw/CNN-SuccSite/ .
    背景与目标: :琥珀酰化是蛋白质翻译后修饰(PTM)的一种,它可以在多种细胞过程中发挥重要作用。由于从高通量质谱法(MS)获得的位点特异性琥珀酰化肽的数量不断增加,已开发出各种工具来计算识别蛋白质上的琥珀酰化位点。但是,这些工具大多数都基于传统的机器学习方法来预测琥珀酰化位点。因此,这项工作旨在基于深度学习模型进行琥珀酰化位点的预测。经过MS验证的琥珀酰化肽的丰度使得可以通过基于序列的属性(例如位置特异性氨基酸组成,k间距氨基酸对的组成(CKSAAP)和位置特异性)研究琥珀酰化位点的底物位点特异性得分矩阵(PSSM)。另外,通过将所有琥珀酰化序列分成具有保守底物基序的几组,采用最大依赖性分解(MDD)来检测赖氨酸琥珀酰化位点的底物特征。根据十倍交叉验证的结果,使用PSSM和信息丰富的CKSAAP属性训练的深度学习模型可以达到最佳的预测性能,并且其性能也要优于传统的机器学习方法。此外,使用了训练数据集中实际上不存在的独立测试数据集,将所提出的方法与六个现有的预测工具进行了比较。测试数据集包括218个阳性实例和2621个阴性实例,所提出的模型可以产生有希望的性能,灵敏度为84.40%,特异性为86.99%,准确度为86.79%,MCC值为0.489。最后,所提出的方法已实现为基于Web的预测工具(CNN-SuccSite),现在可从http://csb.cse.yzu.edu.tw/CNN-SuccSite/免费访问。
  • 【基于机器学习的玻璃化转变温度预测模型:聚羟基链烷酸酯均聚物和共聚物的案例。】 复制标题 收藏 收藏
    DOI:10.1021/acs.jcim.9b00807 复制DOI
    作者列表:Pilania G,Iverson CN,Lookman T,Marrone BL
    BACKGROUND & AIMS: :Polyhydroxyalkanoate-based polymers-being ecofriendly, biosynthesizable, and economically viable and possessing a broad range of tunable properties-are currently being actively pursued as promising alternatives for petroleum-based plastics. The vast chemical complexity accessible within this class of polymers gives rise to challenges in the rational discovery of novel polymer chemistries for specific applications. The burgeoning field of polymer informatics addresses this challenge via providing tools and strategies for accelerated property prediction and materials design via surrogate machine-learning models built on reliable past data. In this contribution, we use glass transition temperature Tg as an example target property to demonstrate promise of the data-enabled route to accelerated learning of accurate structure-property mappings in PHA-based polymers. Our analysis uses a data set of experimentally measured Tg values, polymer molecular weights, and a polydispersity index for PHA-based homo- and copolymers that was carefully assembled from the literature. A fingerprinting scheme that captures key properties based on topology, shape, and charge/polarity of specific chemical units or motifs forming the polymer backbone was devised to numerically represent the polymers. A validated statistical learning model is then developed to allow for a mapping of the polymer fingerprints onto the property space in a physically meaningful and reliable manner. Once developed, the model can not only rapidly predict the property of new PHA polymers but also provide uncertainties underlying the predictions. The model is further combined with an evolutionary-algorithm-based search strategy to efficiently identify multicomponent polymer compositions with a prespecified Tg. While the present contribution is focused specifically on Tg, the surrogate model development approach put forward here is general and can, in principle, be extended to a range of other properties.
    背景与目标: :基于聚羟基链烷酸酯的聚合物具有生态友好性,生物合成性和经济可行性,并且具有广泛的可调性能,目前正积极地用作石油基塑料的有前途的替代品。在这类聚合物中可利用的巨大化学复杂性在合理发现用于特定应用的新型聚合物化学中提出了挑战。高分子信息学的新兴领域通过提供基于可靠的过去数据的替代机器学习模型来提供加速性能预测和材料设计的工具和策略,从而解决了这一挑战。在这一贡献中,我们以玻璃化转变温度Tg作为示例目标特性,以证明基于数据的途径有望加速学习基于PHA的聚合物中精确的结构特性图。我们的分析使用了一组根据实验测得的Tg值,聚合物分子量以及基于文献的PHA基均聚物和共聚物的多分散指数的数据集。设计了一种指纹图谱,该图谱可基于拓扑,形状以及形成聚合物主链的特定化学单元或基序的电荷/极性捕获关键特性,以数字方式表示聚合物。然后开发经过验证的统计学习模型,以允许以物理上有意义和可靠的方式将聚合物指纹映射到属性空间上。一旦建立,该模型不仅可以快速预测新型PHA聚合物的性质,而且还可以提供预测所依据的不确定性。该模型还与基于进化算法的搜索策略结合在一起,可以有效地识别具有预定Tg的多组分聚合物成分。尽管当前的贡献专门针对Tg,但是这里提出的替代模型开发方法是通用的,并且原则上可以扩展到其他属性的范围。
  • 【超连续谱中无赖孤子的机器学习分析。】 复制标题 收藏 收藏
    DOI:10.1038/s41598-020-66308-y 复制DOI
    作者列表:Salmela L,Lapre C,Dudley JM,Genty G
    BACKGROUND & AIMS: :Supercontinuum generation is a highly nonlinear process that exhibits unstable and chaotic characteristics when developing from long pump pulses injected into the anomalous dispersion regime of an optical fiber. A particular feature associated with this regime is the long-tailed "rogue wave"-like statistics of the spectral intensity on the long-wavelength edge of the supercontinuum, linked to the generation of a small number of "rogue solitons" with extreme red-shifts. Whilst the statistical properties of rogue solitons can be conveniently measured in the spectral domain using the real-time dispersive Fourier transform technique, we cannot use this technique to determine any corresponding temporal properties since it only records the spectral intensity and one loses information about the spectral phase. And direct temporal characterization using methods such as the time-lens has resolution of typically 100's of fs, precluding the measurement of solitons which possess typically much shorter durations. Here, we solve this problem by using machine learning. Specifically, we show how supervised learning can train a neural network to predict the peak power, duration, and temporal walk-off with respect to the pump pulse position of solitons at the edge of a supercontinuum spectrum from only the supercontinuum spectral intensity without phase information. Remarkably, the network accurately predicts soliton characteristics for a wide range of scenarios, from the onset of spectral broadening dominated by pure modulation instability to near octave-spanning supercontinuum with distinct rogue solitons.
    背景与目标: :超连续谱的产生是一个高度非线性的过程,当从长脉冲注入到光纤的异常色散状态时,会表现出不稳定和混沌的特性。与该机制相关的一个特殊特征是超连续谱长波长边缘上光谱强度的长尾“流氓波”状统计,与少数带有极端红色的“流氓孤子”的产生有关。转变。尽管可以使用实时色散傅里叶变换技术在光谱域中方便地测量流氓孤子的统计属性,但由于该技术仅记录光谱强度并且会丢失有关光谱的信息,因此我们无法使用此技术来确定任何相应的时间属性阶段。而且,使用诸如时间透镜之类的方法进行的直接时间特征分析,其分辨率通常为fs的100's,这不包括对通常具有更短持续时间的孤子进行测量的方法。在这里,我们通过使用机器学习解决了这个问题。具体来说,我们展示了监督学习如何训练神经网络,以仅从超连续谱光谱强度而不包含相位信息的情况下,预测超连续谱频谱边缘处的孤子的泵浦脉冲位置的峰值功率,持续时间和时间偏移。值得注意的是,该网络可以准确预测各种场景的孤子特性,从以纯调制不稳定性为主导的频谱展宽到具有不同流氓孤子的接近八度的超连续谱。
  • 【假设和数据驱动的多发病率脆弱指数比较:一种机器学习方法。】 复制标题 收藏 收藏
    DOI:10.2196/16213 复制DOI
    作者列表:Peng LN,Hsiao FY,Lee WJ,Huang ST,Chen LK
    BACKGROUND & AIMS: BACKGROUND:Using big data and the theory of cumulative deficits to develop the multimorbidity frailty index (mFI) has become a widely accepted approach in public health and health care services. However, constructing the mFI using the most critical determinants and stratifying different risk groups with dose-response relationships remain major challenges in clinical practice. OBJECTIVE:This study aimed to develop the mFI by using machine learning methods that select variables based on the optimal fitness of the model. In addition, we aimed to further establish 4 entities of risk using a machine learning approach that would achieve the best distinction between groups and demonstrate the dose-response relationship. METHODS:In this study, we used Taiwan's National Health Insurance Research Database to develop a machine learning multimorbidity frailty index (ML-mFI) using the theory of cumulative diseases/deficits of an individual older person. Compared to the conventional mFI, in which the selection of diseases/deficits is based on expert opinion, we adopted the random forest method to select the most influential diseases/deficits that predict adverse outcomes for older people. To ensure that the survival curves showed a dose-response relationship with overlap during the follow-up, we developed the distance index and coverage index, which can be used at any time point to classify the ML-mFI of all subjects into the categories of fit, mild frailty, moderate frailty, and severe frailty. Survival analysis was conducted to evaluate the ability of the ML-mFI to predict adverse outcomes, such as unplanned hospitalizations, intensive care unit (ICU) admissions, and mortality. RESULTS:The final ML-mFI model contained 38 diseases/deficits. Compared with conventional mFI, both indices had similar distribution patterns by age and sex; however, among people aged 65 to 69 years, the mean mFI and ML-mFI were 0.037 (SD 0.048) and 0.0070 (SD 0.0254), respectively. The difference may result from discrepancies in the diseases/deficits selected in the mFI and the ML-mFI. A total of 86,133 subjects aged 65 to 100 years were included in this study and were categorized into 4 groups according to the ML-mFI. Both the Kaplan-Meier survival curves and Cox models showed that the ML-mFI significantly predicted all outcomes of interest, including all-cause mortality, unplanned hospitalizations, and all-cause ICU admissions at 1, 5, and 8 years of follow-up (P<.01). In particular, a dose-response relationship was revealed between the 4 ML-mFI groups and adverse outcomes. CONCLUSIONS:The ML-mFI consists of 38 diseases/deficits that can successfully stratify risk groups associated with all-cause mortality, unplanned hospitalizations, and all-cause ICU admissions in older people, which indicates that precise, patient-centered medical care can be a reality in an aging society.
    背景与目标: 背景:利用大数据和累积赤字理论来发展多发病率脆弱指数(mFI)已成为公共卫生和医疗保健服务中被广泛接受的方法。但是,使用最关键的决定因素构建mFI并通过剂量-反应关系对不同风险组进行分层仍然是临床实践中的主要挑战。
    目的:本研究旨在通过使用机器学习方法来开发mFI,该方法基于模型的最佳适应度选择变量。此外,我们的目标是使用机器学习方法进一步建立4个风险实体,以实现组之间的最佳区分并证明剂量反应关系。
    方法:在这项研究中,我们使用台湾国家健康保险研究数据库,使用单个老年人的累积疾病/赤字理论开发了机器学习多发病率脆弱指数(ML-mFI)。与传统的mFI相比,传统的mFI基于专家意见来选择疾病/缺陷,我们采用随机森林法来选择最有影响力的疾病/缺陷,从而预测老年人的不良后果。为确保生存曲线在随访期间显示出具有重叠的剂量反应关系,我们开发了距离指数和覆盖指数,可在任何时间点将其用于将所有受试者的ML-mFI分为以下几类:适合,轻度虚弱,中度虚弱和严重虚弱。进行生存分析以评估ML-mFI预测不良结局的能力,例如计划外的住院治疗,重症监护病房(ICU)入院率和死亡率。
    结果:最终的ML-mFI模型包含38种疾病/缺陷。与传统的mFI相比,这两个指数在年龄和性别上都有相似的分布模式。然而,在65岁至69岁的人群中,平均mFI和ML-mFI分别为0.037(SD 0.048)和0.0070(SD 0.0254)。差异可能是由mFI和ML-mFI中选择的疾病/缺陷引起的。这项研究共纳入86133名年龄在65至100岁之间的受试者,并根据ML-mFI分为4组。 Kaplan-Meier生存曲线和Cox模型均显示ML-mFI可以显着预测所有有意义的结果,包括随访1年,5年和8年的全因死亡率,计划外住院以及全因ICU入院(P <.01)。特别是,在4个ML-mFI组和不良预后之间发现了剂量反应关系。
    结论:ML-mFI由38种疾病/缺陷组成,可以成功地将与老年人的全因死亡率,计划外住院和全因ICU入院相关的风险组分层,这表明可以进行以患者为中心的精确医疗在老龄化社会中的现实。
  • 【奖励幅度和训练频率对杰克逊港鲨鱼Heterodontus portusjacksoni的学习率和记忆力的影响。】 复制标题 收藏 收藏
    DOI:10.1007/s10071-020-01402-2 复制DOI
    作者列表:Heinrich DDU,Vila Pouca C,Brown C,Huveneers C
    BACKGROUND & AIMS: :The development of adaptive responses to novel situations via learning has been demonstrated in a wide variety of animal taxa. However, knowledge on the learning abilities of one of the oldest extant vertebrate groups, Chondrichthyes, remains limited. With the increasing interest in global wildlife tourism and shark feeding operations, it is important to understand the capacities of these animals to form associations between human activities and food. We used an operant conditioning regime with a simple spatial cognitive task to investigate the effects of reinforcement frequency and reward magnitude on the learning performance and memory retention of Port Jackson sharks (Heterodontus portusjacksoni). Twenty-four Port Jackson sharks were assigned one of four treatments differing in reward magnitude and reinforcement frequency (large magnitude-high frequency; large magnitude-low frequency; small magnitude-high frequency; small magnitude-low frequency). The sharks were trained over a 21-day period to compare the number of days that it took to learn to pass an assigned door to feed. Sharks trained at a high reinforcement frequency demonstrated faster learning rates and a higher number of passes through the correct door at the end of the trials, while reward magnitude had limited effects on learning rate. This suggests that a reduction in reinforcement frequency during tourism-related feeding operations is likely to be more effective in reducing the risk of sharks making associations with food than limiting the amount of food provided.
    背景与目标: :在各种各样的动物分类中,已经证明了通过学习对新情况做出适应性反应的发展。但是,关于现存最古老的脊椎动物群之一的软骨鱼类的学习能力的知识仍然有限。随着对全球野生动植物旅游和鲨鱼饲养活动的兴趣日益浓厚,重要的是要了解这些动物在人类活动与食物之间形成联系的能力。我们使用具有简单空间认知任务的操作条件条件来研究强化频率和奖励幅度对杰克逊港鲨(Heterodontus portusjacksoni)学习成绩和记忆力保持的影响。为二十四只杰克逊港鲨鱼分配了奖励幅度和强化频率不同的四种处理方法之一(大幅度高频率;大幅度低频率;小幅度高频率;小幅度低频)。对鲨鱼进行了为期21天的培训,以比较学习通过指定的门进食所需的天数。在试验结束时,以高加固频率训练的鲨鱼显示出更快的学习速度和通过正确门的次数,而奖励幅度对学习速度的影响有限。这表明,在减少与旅游有关的进食过程中,加固频率的降低可能比限制提供的食物数量更有效地降低了鲨鱼与食物发生关联的风险。
  • 【视障录取标准和学术支持对有学习障碍的大学毕业生的学业成就。】 复制标题 收藏 收藏
    DOI:10.1177/0022219419884064 复制DOI
    作者列表:Sarid M,Meltzer Y,Raveh M
    BACKGROUND & AIMS: :Postsecondary entrance examination scores are generally low predictors of college achievement (grade point average [GPA]) for students with learning disabilities (LD). The difficulties with meeting academic requirements have raised the awareness of the needs of students with LD for support services. The present study examined the adequacy of entrance criteria to academic studies for students with LD and the effectiveness of three support levels during their academic studies in increasing their academic gains. Data were collected for 315 college graduates with LD and 955 graduates who do not have LD (NLD) who completed their BA studies in a college in Israel. Although the admission scores of graduates with LD were lower than those of NLD graduates, their GPA was higher than the GPA of NLD graduates. The psychometric score had a low correlation with final college GPA of graduates with LD. There was no difference in the final GPA of the graduates with LD in the different support programs. The findings suggest that academic support can be an effective way to help students with LD to eliminate or close the gap between them and NLD students and to earn an academic degree, regardless of what if any admission criteria were applied to them.
    背景与目标: :中学入学考试成绩通常是学习障碍学生(LD)的大学成就(平均成绩[GPA])的低指标。满足学业要求的困难提高了人们对LD学生对支持服务的需求的认识。本研究调查了LD学习者入学标准是否充分以及在学习过程中三个支持水平对提高学业成绩的有效性。收集了315名具有LD的大学毕业生和955名没有LD(NLD)的毕业生的数据,这些毕业生在以色列的一所大学完成了学士学位学习。尽管LD毕业生的入学分数低于NLD毕业生的入学分数,但他们的GPA却高于NLD毕业生的GPA。心理测验分数与LD毕业生的最终大学GPA无关。在不同的支持计划中,具有LD的毕业生的最终GPA没有差异。研究结果表明,学术支持可以成为帮助有LD的学生消除或缩小他们与NLD学生之间的鸿沟并获得学位的有效方法,而不管是否适用任何录取标准。
  • 【评估护理学学生的在线解剖学和生理学数字混合学习模式。】 复制标题 收藏 收藏
    DOI:10.1097/CIN.0000000000000639 复制DOI
    作者列表:Barbagallo MS,Porter JE,Lamunu M
    BACKGROUND & AIMS: :Blended online and digital learning includes a variety of activities that combine engaging classroom-based education with online learning. The aim of this study is to evaluate undergraduate students' perceptions of a blended online and digital curriculum for anatomy and physiology in the nursing degree program. A quantitative methodology was used with a sample of 100 undergraduate nursing students from a single Australian University. Descriptive statistics are and presented in this article. Of the 100 participants, 90% were enrolled in the standard mode compared to 10% in a flexible mode of delivery. Results indicated that 29% of participants preferred laboratory classes as having the most impact. Participants (46%) also agreed that recorded lectures were useful, with 36% indicating that the online platform was easy to navigate going as far as wanting more online quizzes (49%). More than half of the participants (54%) acknowledged that anatomy and physiology was important for their future careers. The blended online and digital learner preferences in delivering anatomy and physiology-related courses should be adjusted in order for learning to be effective for undergraduate students in the future.
    背景与目标: :混合的在线和数字学习包括将课堂教学与在线学习相结合的各种活动。这项研究的目的是评估本科生对护理学位课程中有关解剖学和生理学的在线和数字混合课程的看法。定量方法用于来自单个澳大利亚大学的100名本科护理学生的样本。本文将介绍描述性统计信息。在100名参与者中,有90%以标准模式注册,而在灵活交付模式中则为10%。结果表明,有29%的参与者更喜欢实验室课程,因为其影响最大。参与者(46%)也同意录制的讲座非常有用,有36%的参与者表示,在线平台易于浏览,甚至需要更多在线测验(49%)。一半以上的参与者(54%)承认,解剖学和生理学对他们的未来职业很重要。应调整在线和数字学习者在提供解剖学和生理学相关课程方面的偏好,以使学习在将来对本科生有效。
  • 【社交厌恶性泛化学习可以使视皮层神经元适应面部识别提示。】 复制标题 收藏 收藏
    DOI:10.7554/eLife.55204 复制DOI
    作者列表:Stegmann Y,Ahrens L,Pauli P,Keil A,Wieser MJ
    BACKGROUND & AIMS: :Defensive system activation promotes heightened perception of threat signals, and excessive attention to threat signals has been discussed as a contributory factor in the etiology of anxiety disorders. However, a mechanistic account of attentional modulation during fear-relevant processes, especially during fear generalization remains elusive. To test the hypothesis that social fear generalization prompts sharpened tuning in the visuocortical representation of social threat cues, 67 healthy participants underwent differential fear conditioning, followed by a generalization test in which participants viewed faces varying in similarity with the threat-associated face. We found that generalization of social threat sharpens visuocortical tuning of social threat cues, whereas ratings of fearfulness showed generalization, linearly decreasing with decreasing similarity to the threat-associated face. Moreover, individuals who reported greater anxiety in social situations also showed heightened sharpened tuning of visuocortical neurons to facial identity cues, indicating the behavioral relevance of visuocortical tuning during generalization learning.
    背景与目标: :防御系统的激活促进了对威胁信号的高度感知,并且对焦虑信号的过多关注已被讨论为焦虑症病因的一个促成因素。然而,在与恐惧相关的过程中,特别是在恐惧泛化过程中,注意力调节的机制描述仍然难以捉摸。为了检验社会恐惧泛化促进社交威胁提示的视觉皮层表征急剧调整的假设,对67名健康参与者进行了不同的恐惧条件调节,然后进行了泛化测试,参与者在其中观察了与威胁相关面孔相似度不同的面孔。我们发现,社交威胁的泛化使社交威胁提示的视觉皮层调整更加尖锐,而恐惧感的等级显示出泛化,随着与威胁相关面孔的相似度降低,线性下降。此外,在社交场合中表现出更大焦虑的人还显示,大脑皮层神经元对面部识别线索的调节更加尖锐,这表明在泛化学习过程中视觉皮层调节的行为相关性。

+1
+2
100研值 100研值 ¥99课程
检索文献一次
下载文献一次

去下载>

成功解锁2个技能,为你点赞

《SCI写作十大必备语法》
解决你的SCI语法难题!

技能熟练度+1

视频课《玩转文献检索》
让你成为检索达人!

恭喜完成新手挑战

手机微信扫一扫,添加好友领取

免费领《Endnote文献管理工具+教程》

微信扫码, 免费领取

手机登录

获取验证码
登录