Prediction of utilizable true protein of mixed rations for sheep

格式：pdf
大小：596.14 KB
文档页数：12

下载文档原格式

/ 12

蛋白质结构预测中的机器学习方法

蛋白质结构预测中的机器学习方法蛋白质是生命体系中非常重要的分子，因为它们的结构和功能对细胞的正常运作至关重要。

预测蛋白质结构是生物科学领域中的一个重要问题，因为它有助于我们更好地理解蛋白质的生物功能、药物作用等方面。

通过理解蛋白质结构预测中的机器学习方法，我们可以更好地了解这个问题。

在科学家开始研究蛋白质结构预测之前，了解蛋白质结构的基本知识是很有必要的。

蛋白质具有四级结构，包括原生、二级、三级和四级结构。

原生结构是在蛋白质合成过程中形成的。

二级结构是由蛋白质内α-螺旋和β-折叠形成。

三级结构描述的是蛋白质的立体构象，包括螺旋、β-折叠、卷曲和其他结构特征。

最后，四级结构描述的是由多个蛋白质聚合而成的蛋白质复合物结构。

在蛋白质结构预测中，机器学习方法是非常有用的。

机器学习技术是一种通过数据和模型进行预测、分类和决策的方法，而不是基于人工指定规则的方法。

这些方法通过让计算机学习大量的数据，来预测和分类输入数据。

在蛋白质结构预测中，机器学习方法可以帮助我们更好地理解蛋白质结构的模式。

机器学习中的一种常用方法是神经网络。

神经网络是一种通过相互连接的神经单元来模拟人脑神经细胞网络的模型。

在蛋白质结构预测中，神经网络可以用来预测蛋白质的二级结构。

使用神经网络进行二级结构预测的一种流行方法是使用全卷积神经网络，这种网络可以将所有的输入序列转换为输出序列。

另一种机器学习方法是支持向量机（SVM）。

SVM是一种算法，可以将输入数据映射到高维空间中，并在其上构建超平面。

在蛋白质结构预测中，SVM可以用于预测蛋白质的三级结构。

它可以通过提取结构特征来预测蛋白质的空间构型，在进行预测之前，需要对原始蛋白质序列进行处理。

处理过程包括对序列进行特定的编码，并使用特征提取算法，将蛋白质序列的结构信息转换为特征向量。

总的来说，机器学习方法在蛋白质结构预测中是非常重要的。

预测蛋白质结构是一个大型的计算任务，需要消耗大量的计算资源和数据。

蛋白质结构域预测

蛋白质结构域预测蛋白质结构域预测是蛋白质功能注释中的一个重要任务。

蛋白质结构域是指在蛋白质中具有特定结构和功能的连续序列段。

准确地预测蛋白质结构域可以帮助我们理解蛋白质的功能和作用机制，对药物设计和疾病治疗等领域具有重要意义。

随着高通量测序技术的迅猛发展，大量的蛋白质序列数据被积累，蛋白质结构域预测方法也得到了长足的进步。

基于比对的方法是将待预测序列与已知结构域库中的序列进行比对，根据比对结果来判断待预测序列是否含有特定的结构域。

通过这种方法可以预测到已知结构域的序列，但是对于新发现的结构域或者与已知结构域相似度较低的序列预测效果较差。

基于机器学习的方法是利用已知结构域的序列和非结构域的序列作为训练集，通过机器学习算法构建一个预测模型，然后用该模型对待预测序列进行预测。

这种方法可以预测到新发现的结构域，并且可以预测与已知结构域相似度较低的序列。

目前，基于机器学习的方法在蛋白质结构域预测中占据主导地位。

常见的机器学习算法包括SVM（支持向量机）、DT（决策树）、RF（随机森林）等。

这些算法可以通过学习已知结构域的特征和非结构域的特征，来区分结构域和非结构域的序列。

除了机器学习算法，人工神经网络（ANN）也是常用的预测模型。

人工神经网络模型可以建立一个多层的神经网络，通过自我调整权重和阈值参数来计算输入和输出之间的关系。

通过训练样本，可以优化神经网络的参数，使之能够对待预测序列进行准确的预测。

此外，一些新兴的预测方法也逐渐得到应用。

例如，通过整合不同的预测结果进行综合预测。

这种方法可以利用多个预测方法的优势，提高预测的准确性。

同时，一些基于深度学习的方法也逐渐应用于蛋白质结构域预测中。

深度学习利用多层神经网络模型进行特征学习和表征学习，可以从海量的数据中发现隐藏的规律和模式，进一步提高预测效果。

总的来说，蛋白质结构域的准确预测对于研究生命科学和药物设计具有重要意义。

基于比对和机器学习的方法已经取得了显著的进展，通过不断地创新和技术的进步，预测方法将会更加精确和有效。

小分子与蛋白体外相互作用测定方法

小分子与蛋白体外相互作用测定方法⊙编辑：小余生物分子相互作用是一种基本的生命现象，其相互作用研究是现代生命科学研究的重大问题之一。

现在研究相互作用的技术有：平衡透析法、紫外可见吸收光谱、荧光光谱、电化学法等。

随着分子生药研究深入，一些新的检测技术问世，如微量热泳动技术（MST），那么当我们做蛋白相互作用时，面对如此多的技术，该如何让选择？笔者将目前测定小分子与蛋白的体外相互作用，最常用的四种技术：荧光偏振免疫分析法（fluorescence polarization immunoassay，FPIA）、等温量热滴定仪（ITC）、表面离子体共振技术（SPR）等技术的原理、操作、应用范围以及优缺点做一综述，供各位研究者参考。

其中FPIA和ITC是两种经典方法。

1.荧光偏振免疫分析法（fluorescencepolarization immunoassay，FPIA）1.1 原理FPIA是一种定量免疫分析技术，其基本原理是荧光物质经单一平面的蓝偏振光（485nm）照射后，吸收光能跃入激发态，随后回复到基态，并发出单一平面的偏振荧光（525nm）。

偏振荧光强弱程度与荧光分子的大小呈正相关，与其激发时的转动速度呈反相关。

FPIA最适宜检测小至中等分子物质，常用于药物、激素的测定。

1.2 操作1.2.1 首先针对所测蛋白，选择一个阳性药物，对阳性药物进行荧光标记；1.2.2 配制一系列浓度的标准抗原（厂家在试剂盒中提供提供已知剂量的非标记抗原）；1.2.3 用一系列已知浓度的标准抗原与一定量的标记抗原在一定条件下同限量的特异性抗体进行反映；1.2.4 反应达到平衡后，以反应变量作为纵坐标，反应剂量（一系列标准抗原）作为横坐标，做一条曲线，就得到不同浓度的标准抗原对反应变量的竞争抑制曲线；1.2.5 在同等条件下，用待测抗原与一定量的标记抗原与限量的特异性抗体进行反应，并用同样的方法分离待测抗原中的B和F，测量其放射性，计算出B/F，在剂量反应曲线上就可以查出对应的抗原浓度。

信息蛋白质和细胞迁移之间的新发现

信息蛋白质和细胞迁移之间的新发现
佚名
【期刊名称】《基础医学与临床》
【年(卷),期】2005(25)11
【摘要】据美国BIOCOMPARE科技新闻网（2005／9／30）报道，芝加哥伊利诺伊大学的研究人员指出，调控细胞重要信息途径的名为Raf激酶抑制剂蛋白质或简称为RKIP的蛋白质，负责控制激酶的活动，在控制细胞的活动或迁移中扮演重要的角色。

【总页数】1页(P1061-1061)
【关键词】细胞迁移;信息途径;蛋白质;Raf激酶抑制剂;新发;伊利诺伊大学;科技新闻;研究人员;芝加哥;控制
【正文语种】中文
【中图分类】R329.2;R285
【相关文献】
1.环状蛋白质精氨酸甲基转移酶5在结直肠癌细胞增殖、迁移中的作用和机制 [J], 谢津璧;郭润生;王善娟;张丽航
2.蛋白质精氨酸甲基转移酶5促进人卵巢癌HO8910细胞迁移 [J], 赵康容;孙爱琴;林琼;邵根宝
3.新发现抑制某些癌细胞而刺激另一些细胞的蛋白质 [J], 金人一
4.微小RNA-15a-5p靶向蛋白质二硫键异构酶A6前体蛋白抑制前列腺癌细胞增殖、迁移及侵袭 [J], 向从明;陈友干;孙承文;张笑;吴国胜
5.癌细胞中新发现蛋白质可用于研制全新抗癌药物 [J],
因版权原因，仅展示原文概要，查看原文内容请购买。

干扰素刺激基因15抗病毒感染的分子机制

·综述·Chinese Journal of Animal Infectious Diseases中国动物传染病学报摘要：干扰素刺激基因15（ISG15）是由病原微生物或干扰素诱导产生的一种大小为15 kDa 的泛素样蛋白。

在干扰素诱导的数百个干扰素刺激基因中，ISG15是诱导最强烈、最快的ISG 蛋白之一。

研究表明，ISG15对多种病毒具有抗病毒作用。

此外，ISG15在调节宿主损伤、DNA 修复，调节信号通路及抗原递呈中也发挥着重要的作用。

文章介绍了ISG15的概况，并阐述了近年来ISG15在抗病毒、免疫调节和调节宿主信号通路过程中的作用。

关键词：干扰素刺激基因15；抗病毒作用；免疫调节中图分类号：S852.4 文献标志码：A 文章编号：1674-6422（2023）06-0170-07Molecular Mechanism of Interferon-Stimulated Gene 15 Antiviral InfectionTANG Jingyu 1, DU Hanyu 1,2, JIA Nannan 1, TANG Aoxing 1, LIU Chuncao 1, ZHU Jie 1, MENGChunchun 1, LI Chuanfeng 1, LIU Guangqing 1(1. Shanghai V eterinary Research Institute, CAAS, Shanghai 200241, China; 2. Xinjiang Agricultural University, Xinjiang 830052, China)收稿日期：2021-11-02作者简介：国家重点研发计划项目（2016YFD0500108）；中国农业科学院创新工程项目作者简介：唐井玉，女，博士研究生，预防兽医学专业通信作者：刘光清，E-mail：**************.cn干扰素刺激基因15抗病毒感染的分子机制唐井玉1，杜汉宇1,2，贾楠楠1，汤傲星1，刘春草1，朱杰1，孟春春1，李传峰1，刘光清1（1.中国农业科学院上海兽医研究所小动物传染病预防与控制创新团队，上海200241；2.新疆农业大学，乌鲁木齐830052）2023,31(6):170-176Abstract: Interferon-stimulated gene 15 (ISG15) is a ubiquitin-like protein of approximately 15 kDa induced by pathogenic microorganisms or interferons. Among the hundreds of interferon-stimulated genes induced by interferons, ISG15 is one of the most strongly and fastest induced ISG proteins. Studies have shown that ISG15 has antiviral effects against a variety of viruses. In addition, ISG15 plays an important role in regulating host damage, DNA repair, and regulating signaling pathways and antigen delivery. The article presented an overview of ISG15 and described the role of ISG15 in the process of antiviral, immunomodulation and regulation of host signaling pathways in recent years.Key words: Interferon-stimulated gene 15; antiviral infection; immunomodulation先天性免疫应答是抵抗入侵病原体的第一道防线，病原体可以通过宿主模式识别受体来感知。

国家人类基因组北方研究中心

BBRC 1999, 259:569-575
Blue Native PAGE of multiprotein complex from whole cellular lysate
Dialysis permits the analysis of multiprotein complexes of whole cellular lysates by BNPAGE.
GST-mFas fusion proteins
1 149 166 204 293 306
194
306
194
292
194
283
194
276
194
268
194
221
194 221 306
306
GST Pulldown
GST-mFas-associated polypeptides from 32Slabeled HeLa, L929, and Jurkat cell lysates
Blue Native PAGE
SDS-PAGE
Blue Native PAGE of chloroplast thylakoid membranes
BN-PAGE of solubilized chloroplast thylakoid membranes (a) followed by SDS–PAGE in the second dimension (b). CF0F1 ATP synthase was indicated.
Types of Proteomics
Expression Proteomics – Quantitative study of protein expression and their changes between samples that differs by some variable

蛋白质结构预测的方法与工具

蛋白质结构预测的方法与工具蛋白质结构是生物学研究中一个非常重要的领域，因为它对于蛋白质的功能和相互作用有着非常大的影响。

蛋白质结构预测是研究蛋白质学中的一个重要分支，其目的是通过计算机模拟和其他实验手段，预测蛋白质的三维结构。

本文将介绍一些常见的蛋白质结构预测方法和工具。

1. 能量函数蛋白质的三维结构由其氨基酸序列决定。

由于在氨基酸之间的相互作用非常复杂，将其精确地预测出来非常困难。

因此，实际上我们常常用一系列能量函数，来猜测最有可能的三维结构。

能量函数的基本思想是，通过计算预测结构与实验结果的对比来选择最有可能的结构。

能量函数可以预测统计力学方程、物理模型和知识库，用于描述蛋白质的相互作用。

能量函数的选择应当根据具体任务的不同于权衡，其准确度、完备性、计算量和鲁棒性各有不同。

2. 基于机器学习的方法机器学习是指从大量的数据中自动提取出模型，从而能够准确地预测未知数据的特点。

在蛋白质结构预测上，机器学习最成功的是基于神经网络的方法。

基于神经网络的方法，可以学习到从蛋白质的氨基酸序列到三维结构的直接映射，而不需要在蛋白质产生结构时太多的假设。

这种方法有非常高的准确度，并且需要的计算量很少。

3. 蛋白质结构预测工具现在有很多好用的蛋白质结构预测工具可以使用，其中一些工具是公共的，可以在互联网上免费使用。

这些工具使用多种预测方法，如用于序列对齐、模拟、统计建模等，来预测蛋白质的三维结构。

一些常用的工具包括I-TASSER、ROSETTA和PHYRE等。

不同的工具有不同的优缺点，应根据需要进行选择。

其中I-TASSER 最为广泛使用，而ROSETTA则更受科学家们喜爱。

总结：蛋白质结构预测是研究蛋白质学中的一个重要分支，它为我们提供了非常重要的信息，有助于我们更深入地理解生命中的分子结构和功能。

这里我们介绍了一些蛋白质结构预测的方法和工具。

通过不断学习和掌握这些方法和工具，我们将能够更好地运用它们来对现实中的生物学问题进行解决。

药物及生物活性小分子发现与分子设计教材

Actual Positive(TP+FN)
Actual Negative(FP+TN)
合计
Predicted Positive(TP+FP)
Predicted Negative(FN+TN)
TP+FP+FN+TN
真阳性率（召回率）：TPR=TP/ (TP+ FN) 被正确判定的正例占总的正例的比重
• 优势： – 同时预测PK(药代动力学)蛋白和PD(药效动力学)蛋白导致的DDIs；
– 预测模型的生物学原理简单；
– 交叉验证和独立验证中预测精度高，AUC达到0.85；
– 错误的配体-蛋白复合物偶联能被该预测方法最小化；
DDI-CPI
——根据药物-蛋白相互作用组预测药物联合作用
diction
——用于生物活性小分子靶点预测
• 特点：
– 结合2D和3D相似性测量； – 预测可针对五个不同生命体； – 数据集包括280381个小分子与2686个目标相互作用，其中66
％的目标是人类的；
SwissTargetPrediction
——用于生物活性小分子靶点预测
a2 6.568394002 6.417589941 6.662162587 6.386047615 5.880153979 5.78062749 5.757473315 6.427766956 ……
SwissTargetPrediction
——用于生物活性小分子靶点预测
• Precision(精确度)-预测分数曲线：
真阴性率（特异度）：TNR=TN/ (FP+TN) 衡量类别0 的判定能力
精确度：Precision= TP / (TP + FP) 被判定的正例中真正的正例样本的比重

Discovery Studio官方教程--预测蛋白质聚集、水溶性、粘度、可开发性等性质

预测蛋白聚集位点（Protein Aggregation）教程介绍抗体等具有治疗功能的蛋白，如果处于比较高的浓度下，会有发生聚集的趋势。

这会导致抗体的活性下降，并引起免疫反应。

抗体的聚集趋势计算是一种衡量蛋白表面氨基酸聚集倾向性的指标。

具有比较高的聚集趋势得分的位点表明了该区域的氨基酸倾向于发生聚集。

因此这些位点的预测，使得我们可以通过氨基酸定点突变的方法来改造蛋白，增强其稳定性。

本教程使用Calculate Aggregation Scores对一个全长IgG1抗体分子(PDB号为1h2h)进行蛋白聚集位点的预测计算，并分析预测的结果。

本教程涵盖如下内容：●聚集趋势得分的计算●分析蛋白聚集位点聚集趋势得分计算在文件浏览器（Files Explorer）中，展开Samples | Tutorials | Protein Modeling文件夹，双击1hzh.pdb文件。

DS将在一个新的3D窗口中打开该蛋白。

图1 1hzh分子窗口Ctrl+H打开Hierarchy窗口，然后选中Water，点击Delete删除蛋白结构中的结晶水分子。

在工具浏览器（Tools Explorer）中，展开Simulation | Change Forcefield，将Forcefield设为CHARMm Polar H，然后点击Apply Forcefield，这将为蛋白赋上CHARMm Polar H力场。

图2 Apply Forcefield设置界面在工具浏览器（Tools Explorer）中，展开Macromolecules | Predict Protein Aggregation工具面板，点击Calculate Aggregation Scores。

在弹出的参数设置界面中，将Input Typed Protein设为1hzh:1HZH，将Cutoff Radius设为5,7,10。

点击Run运行该任务。

该任务在奔腾4，2Gb内存和2.8GHz的计算机平台上运行大约需要3分钟。

alphafold2 predicted results -回复

alphafold2 predicted results -回复Alphafold2 Predicted Results: Unraveling the Future of Protein FoldingIntroduction:Proteins are essential for numerous biological processes in living organisms. They play a pivotal role in maintaining cell structure, as enzymes catalyzing biochemical reactions, and as signaling molecules. Understanding the three-dimensional structure of proteins is critical for comprehending their functions, interactions, and for developing targeted therapeutic interventions. The prediction accuracy of protein folding has long been a holy grail in the field of biology. However, with the recent breakthroughs in artificial intelligence and deep learning, a game-changing protein structure prediction model called Alphafold2 was introduced.What is Alphafold2?Alphafold2 is an artificial intelligence-based model developed by DeepMind, a subsidiary of Alphabet Inc. It aims to predict the 3D structure of proteins using only its amino acid sequence as input.The model builds on the success of its predecessor, Alphafold, and employs a deep learning algorithm known as a transformer neural network.The Process:The prediction process of Alphafold2 can be divided into three main steps: Tertiary Structure Prediction, Secondary Structure Prediction, and Quaternary Structure Prediction.1. Tertiary Structure Prediction:Tertiary structure refers to the arrangement of amino acid residues in a protein chain to form a 3D structure. Alphafold2 predicts this structure by simulating the folding process using a deep learning algorithm. The model initially maps the amino acid sequence using a multiple sequence alignment technique. The resulting information is then fed into the transformer neural network. The network processes the data and outputs the predicted 3D structure.2. Secondary Structure Prediction:Secondary structure refers to the local folding patterns of protein chains, such as alpha-helices and beta-sheets. Alphafold2 utilizes its transformer neural network to predict the secondary structure by analyzing the patterns in the amino acid sequence. This step is crucial for building the overall tertiary structure prediction.3. Quaternary Structure Prediction:Some proteins consist of multiple subunits, called polypeptide chains, that come together to form functional complexes. Alphafold2 can predict the quaternary structure, i.e., how these subunits interact and assemble. By analyzing the interactions between different protein chains, Alphafold2 provides insights into the overall structure and function of the protein complex.Accuracy and Applications:Alphafold2 has demonstrated remarkable accuracy in predicting protein structures. In the 2020 Critical Assessment of Protein Structure Prediction (CASP14) contest, Alphafold2 outperformed other state-of-the-art methods, achieving unprecedented accuracylevels. Its predictions were found to be in close agreement with experimentally determined structures.The potential applications of Alphafold2 are vast. Accurate protein structure predictions can aid in understanding the causes of various diseases, as many disorders arise due to protein misfolding. Alphafold2's predictions can facilitate drug discovery by providing insights into protein-drug interactions and enabling targeted therapeutic interventions. Additionally, this groundbreaking technology can revolutionize synthetic biology, protein engineering, and biotechnology applications, leading to the development of new enzymes and materials with tailored properties.Ethical Considerations:The advent of Alphafold2 raises ethical concerns. With the ability to predict protein structures rapidly and accurately, there is a risk of malicious use, such as the design of harmful proteins or bioweapons. Proper regulation and control measures need to be in place to prevent misuse and ensure responsible applications of this technology.Conclusion:Alphafold2 represents a significant breakthrough in the field of protein structure prediction. Its ability to predict protein structures accurately and rapidly opens doors to numerous possibilities in biomedicine, biotechnology, and drug discovery. However, with these advancements, ethical considerations and responsible use of this technology become paramount. With further research and development, Alphafold2 has the potential to transform the way we understand and harness the power of proteins.。

1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。

This article was downloaded by:[Zhao, Guang-Yong][Zhao, Guang-Yong]On:27 March 2007Access Details:[subscription number 772651375]Publisher:Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UKArchives of Animal NutritionPublication details, including instructions for authors and subscription information:/smpp/title~content=t713453455Prediction of utilizable true protein of mixed rations forsheep using an in vitro incubation techniqueTo cite this Article:, 'Prediction of utilizable true protein of mixed rations for sheepusing an in vitro incubation technique', Archives of Animal Nutrition, 61:2, 103 - 113xxxx:journal To link to this article: DOI:10.1080/17450390701222972URL:/10.1080/17450390701222972D o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007Prediction of utilizable true protein of mixed rations for sheep using an in vitro incubation techniqueYUAN-XIAO LI &GUANG-YONG ZHAOState Key Laboratory of Animal Nutrition,College of Animal Science and Technology,China Agricultural University,Beijing,PR China(Received 4September 2006;accepted 8January 2007)AbstractThe objective of the present experiment was to study the relationship between in vitro utilizable true protein (uTP)and in vivo -uTP of sheep rations by regression analysis.A further aim was to analyse if in vivo -uTP of mixed rations could be predicted by regression analysis between in vitro -uTP and in vivo -uTP,using N-retention of sheep as important evaluation criteria of protein value.Three adult male sheep (body weight [BW]46+1.3kg)ﬁtted with rumen cannulas and simple T-type duodenal cannulas were fed with twelve typical rations with graded levels of crude protein and true protein in four experiments according a 363Latin square design.Each experimental period included an adaptation (7days),a N balance trial (4days)and a collection of duodenal digesta (3days).During collection of duodenal digesta,polyethylene glycol and chromium oxide were used as dual markers for the measurement of duodenal digesta ﬂow and calculation of the in vivo -uTP of duodenal digesta.The in vitro -uTP of the rations was determined using the in vitro incubation technique of Zhao and Lebzien (2000).It was found that both in vitro -uTP intake and in vivo -uTP intake were signiﬁcantly correlated with N-retention (p 50.001)and that there was a signiﬁcant linear relationship between the content of in vitro -uTP and in vivo -uTP in rations (p 50.001).Therefore,it was concluded that the used in vitro incubation technique is suitable for the determination of in vitro -uTP of mixed rations for sheep,and that the amount of in vivo -uTP can be predicted by regression between in vitro -uTP and in vivo -uTP.Keywords:Utilizable true protein,sheep,in vitro incubation,prediction1.IntroductionThe crude protein (CP)ﬂow to the small intestine of ruminants mainly includes rumen undegradable crude protein (UDP)and rumen microbial crude protein (MCP).Feeding ruminants with rumen-undegradable dietary protein increases dietary protein supply to the small intestine but decreases rumen microbial protein synthesis,resulting in only minor differences in total protein supply to small intestine (Clark et al.1992).Therefore,deter-mination of utilizable crude protein (uCP)that includes UDP and MCP is more practical and accurate than separate determination of UDP and MCP in evaluating dietary protein value forCorrespondence:Prof.Guang-Yong Zhao,State Key Laboratory of Animal Nutrition,College of Animal Science and Technology,China Agricultural University,No 2.Yuanmingyuan West Road,Beijing,PR China,100094.E-mail:zhaogy@ Archives of Animal Nutrition April 2007;61(2):103–113ISSN 1745-039X print/ISSN 1477-2817online ª2007Taylor &Francis DOI:10.1080/17450390701222972D o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007ruminants (Lebzien et al.1996).Based on the uCP concept,Zhao and Lebzien (2000)developed an in vitro incubation technique for estimating uCP of feedstuffs for ruminants.They found that in vitro -uCP was highly correlated with in vivo -uCP,estimated by in vivo experimental data.The in vitro incubation technique was later used for the estimation of utilizable amino acids (uAA)(sum of amino acids from undegraded feed protein and microbial protein)of feedstuffs.It was found that there was a signiﬁcant regressive relationship between in vitro -uAA and in vivo -uCP estimated by in vivo experimental data (Zhao &Lebzien 2002).Although estimation of uAA is more accurate than uCP,it is more expensive and complicated.The utilizable true protein (uTP),which includes true feed protein undegradable in rumen and rumen microbial true protein,is more accurate than uCP,and can be determined by simpler and less expensive means than uAA.Since the cost for amino acid analysis is much higher than for uCP,and amino acid analysers are not always available in many laboratories,uTP may be more practically applicable in evaluation of nutritive value of feedstuffs.The in vitro incubation technique of Zhao and Lebzien (2000)was successfully used for the determination of uCP and uAA of feedstuffs,it could be anticipated that the technique could be also used for the determination of uTP.The objective of the present experiment was to study the relationship between in vitro -uTP and in vivo -uTP of sheep rations by regression analysis.A further aim was to analyse whether in vivo -uTP of mixed rations could be predicted by regression analysis between in vitro -uTP and in vivo -uTP,using N-retention of sheep as an important evaluation criteria of protein value.2.Materials and methods 2.1.In vivo experiments2.1.1.Animals and feeding.Three adult crossbred sheep (Small Tailed Han sheep 6Dorset sheep),weighing 46+1.3kg,ﬁtted with rumen cannulas and T-type duodenal cannulas,were used as experimental animals.The animals were kept in individual metabolic cages and fed twice daily at 8:00and 16:00h,respectively.The amount of dry matter given to each animal was 2.6%of BW.Fresh drinking water was freely available.2.1.2.Experimental design.Based on the nutrient content of feedstuffs and the nutrient requirements of ﬁnishing lambs (BW 45kg,BWG 250g/d)12typical rations for sheep with graded levels of CP (7.47–12.89%CP)were formulated according to the National Research Council (NRC 1975).Apart from CP,all other nutrients of the rations were the same and met the nutrient requirement of the sheep.Ingredients and nutrient composition of the rations are shown in Table I.In four experiments,to rations and sheep were allocated to the experimental groups according a 363Latin square design.Three rations were used in each experiment.Each experimental period included 7days of adaptation,4days of N balance trial and a collection of duodenal collection (3days).For the measurement of duodenal digesta ﬂow polyethylene glycol (PEG,molecular weight 4000)and chromium oxide (Cr 2O 3)were used as liquid-phase and solid-phase markers,respectively.2.1.3.Sample collection and measurement.During the N balance trial,urine was collected with the aid of harnesses connected to containers containing 20ml of H 2SO 4(10%,v/v)to keep the pH of urine below 3.0.Faeces were collected in plastic bags.The volume of urine and104Y.-X.Li &G.-Y.ZhaoD o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007weight of faeces were recorded and 5%of urine and 10%of faeces were taken as samples.The samples were kept at 7208C for analysis.After the adaptation period,daily 3.20g PEG and 6.40g Cr 2O 3were mixed well with the diet of each animal.The spot-sampling of duodenal digesta was carried out every 3h starting at 8:30h on the ﬁrst day of digesta collection.Each spot sample of digesta was 20ml.The digesta samples from each sheep were well mixed and homogenized.About 2/3of each pooled sample were kept at 7408C for freeze-drying,and 1/3of the sample was centrifuged at 3000g for 10min for fractionation of liquid-phase and solid-phase sub-samples.The sub-samples were then stored until analysis at 7208C.2.2.In vitro incubation2.2.1.Animals and feeding.Three adult crossbred male sheep (Small Tailed Han sheep 6Dorset sheep),weighing about 42+1.8kg,ﬁtted with rumen cannulas,were used as donors of rumen ﬂuid.The daily ration for the sheep consisted of 980g wild rye hay and 420g concentrate mixture.The concentrate mixture included 66.7%corn,6.7%wheat feed ﬂour,15.3%soybean meal,8.3%cottonseed meal and 3%NaCl.The ration was given to the sheep in two equal meals,at 8:00and 16:00h,respectively,and fresh drinking water was freely available.The 12rations in the in vivo experiment were used as experimental samples.The samples were air-dried and milled through a 3-mm sieve for determination and analysis.2.2.2.Incubation.The in vitro incubation technique of Zhao and Lebzien (2000)was used for the determination of uTP of the rations.The rumen ﬂuid was taken before morning feeding and was strained through four layers of surgical gauze.A volume of 312.5ml rumen ﬂuid from each sheep was used for making-up of rumen ﬂuid/buffer medium mixture.The ﬂask that contained rumen ﬂuid/buffer medium mixture was continuously gassed with CO 2.Approximately 0.5g of feed sample was weighed into each incubation tube.Each of the 12rations was incubated with rumen ﬂuid from every sheep in ten replicates in order to get enough incubation residues for analysis.Ten blank tubes without feed samples were also prepared for each set of samples.About 50ml of rumen ﬂuid/buffer medium mixture wasTable position of experimental rations [%as fed].Ration Wild rye hay Alfalfa hay Corn Soybean meal Cottonseed meal Dicalcium phosphate Calcium carbonate Sodium chloride 170.000.0028.000.000.000.800.20 1.00250.000.0047.900.100.000.600.40 1.00345.00 5.0047.700.500.000.600.20 1.00445.00 5.0046.10 2.100.000.600.20 1.00540.0010.0046.00 2.500.000.500.00 1.00635.0010.0049.500.30 3.700.400.10 1.00735.0010.0047.80 2.60 3.000.400.20 1.00835.0010.0046.20 5.00 2.400.300.10 1.00935.0010.0044.607.20 1.700.400.10 1.001035.0010.0043.009.60 1.000.300.10 1.001130.0015.0042.800.2010.800.100.10 1.001230.0015.0041.002.4010.000.000.101.00Utilizable true protein in sheep rations105D o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007taken to each incubation tube.The tubes were incubated in a water bath of 388C for 24h.The pH was immediately measured at the end of incubation.The incubation residues from the tubes using rumen ﬂuid from the same sheep were mixed.The incubation residues of the blanks were also treated in the same way.All the samples were kept at 7408C for later use.2.3.Determinations and chemical analysisFor DM determination all feed samples were dried for 6h at 1058C.The digesta samples and the in vitro incubation residues were freeze-dried (Freezone 6,Labconco,USA)and milled through a 1-mm sieve.The CP of rations,faecal samples,urinary samples,freeze-dried incubation samples and freeze-dried duodenal samples were determined using the Kjeldahl method.The TP of feed samples,freeze-dried incubation samples and freeze-dried duodenal samples was determined using the tungstic acid method (Licitra et al.1996).A 10%sodium tungstate (Na 2WO 4Á2H 2O)water solution (0.3M)and 0.5M H 2SO 4were made up,respectively.Approximately 0.5g sample was weighed into a 125ml Erlenmeyer ﬂask and then 50ml cold distilled water and 8ml of 10%sodium tungstate solution were added,respectively.The sample and the solution were well mixed.The ﬂasks were left stand overnight at room temperature before ﬁltering.The ﬁlter paper together with the residues was then used for nitrogen determination using the Kjeldahl method.The PEG content of liquid-phase sub-samples and solid-phase sub-samples of duodenal digesta was analysed using the method of Malawer and Powell (1967),and the Cr 2O 3content of sub-samples was analysed by the modiﬁed wet digestion method of Stevenson and Langen (1960).2.4.Calculation and statistical analysisThe duodenal DM ﬂows were calculated according to the double-phase marker technique of Faichney (1975):x ÁS D þy ÁS F ¼x ÁP D þy ÁP F y x ¼P D ÀS D S F ÀP F¼R :ð1ÞWhere x ¼a quantity of digesta (D ),[kg];y ¼a quantity of ﬂuid (F )which,when added to or removed from x ,reconstitutes true digesta (TD ),[kg];S D ¼Fractional concentration of soluble marker (PEG)in solid-phase,[g Ákg 71]Á[g 71];S F ¼Fractional concentration of soluble marker (PEG)in liquid-phase,[g Ákg 71]Á[g 71];S TD ¼Fractional concentration of soluble marker (PEG)in true digesta (TD ),[g Ákg 71]Á[g 71];P D ¼Fractional concentration of particle marker (Cr 2O 3)in solid-phase,[g Ákg 71]Á[g 71];P F ¼Fractional concentration of particle marker (Cr 2O 3)in liquid-phase,[g Ákg 71]Á[g 71];P TD ¼Fractional concentration of particle marker (Cr 2O 3)in true digesta (TD ),[g Ákg 71]Á[g 71];where R is the reconstitution factor,i.e.the number of units of ﬂuid that must be added to (or removed from)one unit of digesta to obtain true digesta.ThenS D þR ÁS F 1þR ¼S TD ¼P D þR ÁP F1þR ¼P TDð2ÞTD flow ¼1=S TD ¼1=P TDð3Þ106Y.-X.Li &G.-Y.ZhaoD o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007when S D ,S F ,P D and P F of duodenal contents were calculated,R could be calculated using Equation 1,then S TD or P TD was calculated using Equation 2and TD was calculated using Equation 3.The in vivo -uTP of duodenal digesta was calculated as:In vivo -uTP ¼TP ÁTD =DMIð4Þwhere:In vivo -uTP ¼in vivo utilizable true protein determined [%DMI];TP ¼True protein of freeze-dried duodenal digesta [%DM];TD ¼True duodenal digesta [g DM];and DMI ¼Dry matter intake [g DM].The in vitro -uTP of rations was calculated as:In vitro -uTP ¼ðTP Sample ÁW Sample ÀTP Blank ÁW Blank Þ=DMIncubationð5Þwhere:In vitro -uTP ¼utilizable true protein determined by in vitro incubation technique [%DM];TP Sample ¼TP of incubation residue of ration [%DM];W Sample ¼Weight of incubation residue of ration [g];TP Blank ¼TP of incubation residue of blanks [%DM];W Blank ¼Weight of incubation residue of blanks [g];and DM Incubation ¼Dry matter of feed samples for incubation [g].The N-retention was calculated as:N-retention ½g Ád À1 ¼N-intake ½g Ád À1 ÀFaecal N ½g Ád À1 ÀUrinary N ½g Ád À1 :ð6ÞThe regression analysis and t -test of the results were performed according to SPSS 10.0for Windows.3.ResultsAfter 24h in vitro incubation,the pH value in all incubation tubes were within the range of 6.9–7.1,indicating that the in vitro incubations were similar to normal rumen fermentation.parison of the contents of the CP,TP,in vitro -uTP and in vivo -uTP of the rations The DM,CP,TP,in vivo -uTP and in vitro -uTP of the rations are shown in Table II.The results indicated that TP increased with CP and both in vitro -uTP and in vivo -uTP increased with CP and TP of the rations.Statistical analysis indicated that TP was signiﬁcantly higher than in vivo -uTP (p 50.01)and signiﬁcantly lower than in vitro -uTP and CP (p 50.01).It was also found that in vitro -uTP was signiﬁcantly higher than in vivo -uTP (p 50.01).3.2.Relationships between the CP,TP,in vitro -uTP and in vivo -uTP of the rationsThe regression analysis indicated that there was a signiﬁcant linear relationship between in vitro -uTP [x,%DM]and in vivo -uTP [y,%DM](see Figure 1).There were also a signiﬁcant linear relationships between CP [x,%DM]and in vitro -uTP [y,%DM],and between TP [x,%DM]and in vitro -uTP [y,%DM](see Figure 2).Furthermore,the linear relationships between CP [x,%DM]and in vivo -uTP [y,%DM]and between TP [x,%DM]and in vivo -uTP [y,%DM]were also signiﬁcant (see Figure 3).Utilizable true protein in sheep rations 107D o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007The results indicated that the regression coefﬁcient between in vitro -uTP and in vivo -uTP (r 2¼0.87)was higher than those between CP and in vivo -uTP (r 2¼0.77)and between TP and in vivo -uTP (r 2¼0.78).Both CP and TP of the rations had much closer correlations with in vitro -uTP (r 2¼0.93,r 2¼0.94,respectively)than that with in vivo -uTP (r 2¼0.77,r 2¼0.78,respectively).In vitro -uTP and in vivo -uTP had slightly closer correlation with TP (r 2¼0.94,r 2¼0.78,respectively)than that with CP (r 2¼0.93,r 2¼0.77,respectively).3.3.Relationships between N-retention and intakes of the CP,TP,in vitro -uTP and in vivo -uTP N-retention and intakes of CP,TP,in vitro -uTP and in vivo -uTP of animals are shown in Table III and Table IV,respectively.Table II.Contents of dry matter (DM),crude protein (CP),true protein (TP),in vitro utilizable true protein (in vitro -uTP)and in vivo utilizable true protein (in vivo -uTP)in experimental rations.Ration DM [%]CP [%of DM]TP [%of DM]In vitro -uTP [%of DM]In vivo -uTP [%of DM]190.697.47 5.747.31+0.54 4.70+0.55290.437.78 6.347.12+0.66 4.42+0.16390.488.38 6.829.08+0.37 5.70+0.38490.499.027.349.07+1.24 5.25+0.38590.539.637.829.52+0.66 5.76+0.44690.529.858.1910.27+0.57 5.87+0.65790.5410.558.7511.50+0.66 5.94+0.45890.5311.369.3911.67+0.68 6.11+0.67990.5412.039.9212.05+0.31 6.67+0.861090.5412.7910.5411.93+0.24 6.19+0.501190.6912.2110.3012.52+0.497.03+0.361290.2412.8910.8413.65+0.268.18+0.49Mean90.52+0.0310.33+0.56a8.50+0.50b10.47+0.60a5.98+0.29ca,b,cDifferent superscripts indicate a signiﬁcant difference at p 50.01.Figure 1.Relationship between in vitro utilizable true protein (in vitro -uTP)and in vivo utilizable true protein (in vivo -uTP).108Y.-X.Li &G.-Y.ZhaoD o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007Statistical analysis indicated that there were signiﬁcant regression relationships between N-retention [y,g Ád 71]and CP intake [x,g Ád 71],TP intake [x,g Ád 71],in vitro -uTP intake [x,g Ád 71]in vivo -uTP intake [x,g Ád 71](see Table V).The regression coefﬁcients (r 2)Figure 2.Relationship between nitrogenous components and in vitro utilizable true protein (in vitro-uTP).Figure 3.Relationship between nitrogenous components and in vivo utilizable true protein (in vivo -uTP).Utilizable true protein in sheep rations 109D o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007between N-retention and CP intake and that between N-retention and TP intake were similar.The regression coefﬁcients (r 2)between N-retention and in vitro -uTP intake and that between N-retention and in vivo -uTP intake were also similar,whereas these were higher thanTable III.N-intake,N-excretion and N-retention during the N-balance period.Ration N intake [g Ád 71]Faecal N [g Ád 71]Urinary N [g Ád 71]N retention [g Ád 71]113.73+0.720.69+0.04 3.90+0.629.14+1.06215.00+0.750.71+0.08 5.30+0.968.99+0.67315.37+0.810.76+0.08 5.30+0.379.32+0.99417.41+0.870.69+0.04 6.54+1.7010.18+1.07518.61+0.930.64+0.137.76+1.5210.20+0.89619.01+0.950.78+0.087.82+0.6310.41+0.54720.38+1.020.74+0.087.66+0.9411.99+0.06821.93+1.100.72+0.1110.27+1.3610.94+0.86923.23+1.160.69+0.1210.83+0.7811.71+0.291024.71+1.240.61+0.0713.67+1.4910.43+0.651123.62+1.180.74+0.1012.20+1.0010.68+0.121224.81+1.240.75+0.0910.97+3.1613.09+1.29Table IV.Intake of dry matter (DM),crude protein (CP),true protein (TP),in vitro utilizable true protein (in vitro -uTP)and in vivo utilizable true protein (in vivo -uTP).Ration DM intake [g Ád 71]CP intake [g DM Ád 71]TP intake [g DM Ád 71]In vitro -uTP intake [g DM Ád 71]In vivo -uTP intake [g DM Ád 71]11149+60.585.8+4.5265.9+3.4783.4+2.4453.4+5.0021206+60.393.8+4.6976.4+3.8285.2+5.5753.1+5.1131146+60.396.1+5.0678.2+4.11103.6+1.0761.6+4.4441207+60.3108.8+5.4488.6+4.43108.1+10.6563.0+2.3051207+60.4116.3+5.8294.4+4.72114.1+1.9870.2+2.2961207+60.3118.8+5.9498.8+4.94123.3+1.4870.3+3.7371207+60.4127.4+6.37105.6+5.28138.6+9.2971.2+5.2981207+60.3137.1+6.85113.3+5.67140.1+3.0672.7+5.7091207+60.4145.2+7.26119.8+5.99145.4+7.8380.1+10.47101207+60.4154.4+7.72127.2+6.36143.8+5.6973.4+8.84111209+60.5147.6+7.38124.5+6.23150.8+3.5883.2+6.89121203+60.2155.0+7.75130.4+6.52164.0+7.7097.3+5.24Table V.Result of regression analysis between N-retention [y,g DM Ád 71]and intakes of crude protein (CP),true protein (TP),in vitro utilizable true protein (in vitro -uTP)and in vivo utilizable true protein (in vivo -uTP)[x,g DM Ád 71].Intake of Equation r 2n p CP y ¼0.04x þ5.720.621250.01TPy ¼0.04x þ6.050.631250.01In vitro -uTP y ¼0.04x þ5.510.771250.001In vivo -uTPy ¼0.09x þ4.520.781250.001110Y.-X.Li &G.-Y.ZhaoD o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007the regression coefﬁcient between N-retention and CP intake and that between N-retention and TP intake.4.Discussionparison of in vitro -uTP and in vivo -uTPBy deﬁnition,uTP consists of rumen undegradable true protein and microbial true protein.Both dietary CP and TP could contribute to rumen undegradable true protein and microbial true protein,therefore both in vitro -uTP and in vivo -uTP increased linearly with the contents of CP and TP of the rations.The results were in agreement with those that total non-ammonia N ﬂow from rumen to lower digestive tract increased with dietary CP in dairy cows (Cunningham et al.1996;Reynal et al.2003;Olmos Colmenero &Broderick 2006).From Table II,it could be seen that in vitro -uTP is signiﬁcantly higher than in vivo -uTP (p 50.01).The results could be explained by the following reasons.Firstly,in vivo digestion and in vitro incubation have differences as well as similarities.In in vivo digestion,dietary protein can be extensively degraded and transformed into microbial protein in the rumen,and the degraded products,such as ammonia,can be partly absorbed into blood across rumen wall.Furthermore,protein ﬂowing from the rumen can be partly digested in abomasum and some amino acids and peptides are absorbed before arriving at the duodenum (Webb et al.1992).During in vitro incubation,dietary protein can be degraded and transformed into microbial protein,but in incubation vessels the degraded products cannot be absorbed.Secondly,in vitro incubation,ammonia carbonate was added to buffer as N source for incubation.The added N source might be partly utilized by rumen microorganisms for microbial protein synthesis in incubation.That is,in vivo -uTP is the product of multi-stomach digestion while in vitro -uTP is actually the product of in vitro incubation.Therefore,in vitro -uTP is higher than in vivo -uTP.parison of the regression relationships between CP,TP,in vitro -uTP and in vivo -uTP In vitro incubation technique has been widely used in evaluation of feedstuffs.Many studies indicated that there were close correlations between indices of rumen fermentation and those of in vitro incubation.In the present study,it was found that there was a signiﬁcant linear regression relationship between in vitro -uTP and in vivo -uTP (r 2¼0.87,p 50.001).The results indicated the effectiveness of in vitro incubation technique in evaluation of feedstuffs when using uTP as a criterion.The regression coefﬁcient between CP and in vivo -uTP (r 2¼0.77)and that between TP and in vivo -uTP (r 2¼0.78)were lower than that between in vitro -uTP and i n vivo -uTP (r 2¼0.87).The results indicated that for predicting in vivo -uTP of the rations in vitro -uTP was more accurate than CP and TP.The reason for this could be that CP and TP were based on chemical analysis and rumen fermentation was not reﬂected in the indices,therefore the precision was less than that of in vitro -uTP.Results also showed that both CP and TP of the rations had a much closer correlation with in vitro -uTP (r 2¼0.93,r 2¼0.94,respectively)than with in vivo -uTP (r 2¼0.77,r 2¼0.78,respectively).This could be explained by the differences between in vivo digestion and in vitro digestion as discussed in 4.1.4.3.Factors affecting the measurement of in vivo -uTPNormally non-ammonia N ﬂow at the duodenum of ruminants is higher than N intake when dietary CP is low,because rumen microbial protein synthesis from recycled N andUtilizable true protein in sheep rations 111D o w n l o a d e d B y : [Z h a o , G u a n g -Y o n g ] A t : 07:03 27 M a r c h 2007 endogenous N is increased (Kluth et al.2000).However,it was found that in vivo -uTP content is signiﬁcantly lower than CP content of the rations (p 50.01,see Table II).This might have resulted from several reasons.Firstly,in the present experiment,the duodenal digesta ﬂow was determined using dual markers,i.e.PEG and Cr 2O 3,whereas Kluth et al.(2000)determined the duodenal digesta using TiO 2as a marker.Different marker techniques might result in some differences.Secondly,the rations in the present experiments consisted of limited kinds of feedstuffs,and premix was not included.Therefore,some nutrients such as minerals,trace elements and fat-soluble vitamins in the rations might not be well-balanced for the growth and reproduction of rumen microorganisms,thus resulting in low microbial protein synthesis even if recycled-N and endogenous N might be parison of CP,TP,in vitro -uTP and in vivo -uTP of mixed rations for the prediction of N-retention In the present experiment the results indicated that N-retention was positively correlated with the intakes of CP,TP,in vitro -uTP and in vivo -uTP,and that N-retention was signiﬁcantly inﬂuenced by protein intake.However,different indices,i.e.CP,TP,in vitro -uTP and in vivo -uTP,varied in precision.This could be explained by the regression coefﬁcients (r 2)between N-retention and intakes of CP,TP,in vitro -uTP and in vivo -uTP,which are 0.62,0.63,0.77and 0.78,respectively.The results indicated that in vitro -uTP and in vivo -uTP were more accurate than CP and TP,whereas in vivo -uTP was only numerically more accurate than in vitro -uTP for evaluating protein value of feedstuffs for ruminants.The resultshave indicated in vitro -uTP as a useful index of dietary protein and in vivo -uTP is as accurate as the in vitro -uTP.Due to the difﬁculty of the determination of in vivo -uTP,it was necessary to use in vitro -uTP for the prediction of in vivo -uTP based on the regression relationship between these two indices.5.ConclusionsIt can be concluded that in mixed rations for sheep the amount of in vitro -uTP determined by in vitro incubation technique was closely correlated with the amount of in vivo -uTP.Furthermore,the results have shown that the in vitro incubation technique of Zhao and Lebzien (2000)is suitable for estimation of in vitro -uTP of feedstuffs.Therefore,the in vivo -uTP of mixed rations can be predicted from the in vitro -uTP based on the regression relationship between in vitro -uTP and in vivo -uTP.ReferencesClark JH,Klusmeyer TH,Cameron MR.1992.Microbial protein synthesis and ﬂows of nitrogen fractions to theduodenum of dairy cows.J Dairy Sci 75:2304–2323.Cunningham KD,Cecava MJ,Johnson TR,Ludden PA.1996.Inﬂuence of source and amount of dietary protein onmilk yield by cows in early lactation.J Dairy Sci 79:620–630.Faichney GJ.1975.The use of markers to partition digestion within the gastro-intestinal tract of ruminants.In:McDonald IW,Warner ACI,editors.Digestion and metabolism in the ruminant.The University of New England Publishing Unit.pp 271–291.Kluth H,Gabel M,Voigt J,Scho ¨nhusen U.2000.The use of endogenous nitrogen for microbial crude proteinsynthesis in the rumen of growing bulls.J Anim Physiol Anim Nutr 84:136–147.Lebzien P,Voigt J,Gabel M,Ga ¨deken D.1996.Zur Scha ¨tzung der Menge an nutzbarem Rohprotein am Duodenumvon Milchku ¨hen.J Anim Physiol Anim Nutr 76:218–223.Licitra G,Hernandez TM,Van Soest PJ.1996.Standardization of procedures for nitrogen fractionation of ruminantfeeds.Anim Feed Sci Technol 57:347–358.112Y.-X.Li &G.-Y.Zhao。