本文關(guān)鍵詞：甲型流感病毒H3抗原進(jìn)化及變異規(guī)律研究　出處：《復(fù)旦大學(xué)》2005年博士論文　論文類(lèi)型：學(xué)位論文

  更多相關(guān)文章： 甲型流感病毒 H3亞型 抗原變異 生物信息學(xué) 數(shù)據(jù)挖掘 兩步聚類(lèi) 自組織圖 隨機(jī)森林 進(jìn)化樹(shù) Bayes方法 正向選擇位點(diǎn)

【摘要】：甲型流感病毒長(zhǎng)期以來(lái)一直是威脅人類(lèi)健康的重要病原體之一,對(duì)其變異規(guī)律進(jìn)行深入的分析研究在流感的防治工作中具有重要的現(xiàn)實(shí)意義。本研究借助近年來(lái)發(fā)展起來(lái)的數(shù)據(jù)挖掘技術(shù)和進(jìn)化樹(shù)分析方法,從生物信息學(xué)的角度對(duì)流感病毒H3抗原的變異規(guī)律進(jìn)行了系統(tǒng)研究,主要包括以下內(nèi)容: 1.針對(duì)生物信息數(shù)據(jù)動(dòng)態(tài)更新,不斷增加的特點(diǎn),確定采用數(shù)據(jù)拆分技術(shù)進(jìn)行樣本采集,并使用聚類(lèi)分析對(duì)序列的變異規(guī)律進(jìn)行初步分析。兩步聚類(lèi)和自組織圖這兩種新的聚類(lèi)方法被加以應(yīng)用,結(jié)果顯示全部H3A1序列可以被清楚的分為6類(lèi),各類(lèi)在宿主間的分布特征清楚的反映了不同宿主序列間親緣關(guān)系的遠(yuǎn)近,而時(shí)間分布上則反映了H3亞型病毒在人間不斷傳播和變異的過(guò)程。 2.在聚類(lèi)分析基礎(chǔ)上,使用隨機(jī)森林模型進(jìn)行了人H3序列變異關(guān)鍵位點(diǎn)的分析。借助于隨機(jī)重排技術(shù),最終篩選出了27個(gè)高度變異位點(diǎn),除2號(hào)位點(diǎn)外,其余26個(gè)均屬于五個(gè)抗原決定區(qū)域之一。這一結(jié)果充分說(shuō)明H3A1序列在人間的變異主要應(yīng)當(dāng)是宿主的免疫屏障所驅(qū)動(dòng)的。在變異的預(yù)測(cè)上,這些位點(diǎn)也顯示出了一定的作用。 3.為了進(jìn)一步刻畫(huà)H3序列的變異規(guī)律,本研究繪制出了全部序列的進(jìn)化樹(shù)。分析結(jié)果顯示進(jìn)化樹(shù)結(jié)構(gòu)和聚類(lèi)結(jié)果高度一致,相應(yīng)結(jié)果不僅詳細(xì)刻畫(huà)了數(shù)十年來(lái)流感病毒變異的基本規(guī)律,還明確回答了以下兩個(gè)關(guān)鍵問(wèn)題:豬宿主在人流感病毒變異中所起的作用是非常小的,主要起抗原存儲(chǔ)器的作用;而與全球其余地區(qū)相比,中國(guó)南部地區(qū)在人H3亞型的抗原漂移中并未顯出更高的重要性,不應(yīng)當(dāng)被認(rèn)為是人群中已有亞型新變異株的發(fā)源地。 4.基于已經(jīng)得到的進(jìn)化樹(shù)結(jié)構(gòu),本研究又進(jìn)一步分析了進(jìn)化樹(shù)主干上正向選擇位點(diǎn)的出現(xiàn)規(guī)律,Yang的正向選擇模型被用于分析。結(jié)果表明人流感病毒H3抗原所承受的選擇壓力是逐漸增大的,并且在1995年后達(dá)到高峰。在被篩選出的28個(gè)正向位點(diǎn)中,絕大多數(shù)都只是在流感病毒的某一段進(jìn)化時(shí)期內(nèi)承受著正向選擇壓力。 綜合上述分析結(jié)果,我們認(rèn)為H3亞型流感病毒在人間的變異呈現(xiàn)出逐漸加速的趨勢(shì),而這主要是受到了越來(lái)越強(qiáng)大的免疫屏障的篩選作用所致。目前免疫策略的核心任務(wù)是保護(hù)高危人群,但是這種策略并不能夠阻止,反而可能加速了新變異株的出現(xiàn)。通過(guò)更進(jìn)一步的生物信息學(xué)分析,并且在全球采取統(tǒng)一的主動(dòng)
[Abstract]:Influenza A virus has long been one of the important pathogens that threaten human health, in-depth analysis and Research on the prevention of influenza has important practical significance for its variation. This study uses recently developed data mining technique and phylogenetic analysis method, variation from the perspective of bioinformatics on influenza virus H3 antigen was studied, mainly include the following contents:
1. according to the dynamic biological data update features increasing, determine the sampling data by splitting technique, and use clustering analysis of sequence variation were analyzed. The two step clustering and self organizing map of the two new clustering methods are applied, results showed that all H3A1 sequences can be clearly divided 6 types, distribution characteristics of different kinds in the host between clearly reflected the genetic relationship between different host sequence distance and time distribution is the reflection of the H3 virus subtype in human spread and variation.
2. in the clustering analysis based on the analysis of H3 sequence variation in key sites using random forest model. With the help of random permutation, finally we selected 27 highly variable sites, except for site 2, the remaining 26 belong to the area of one of the five antigenic determinants. This result shows that the sequence of H3A1 in the variation of human should be mainly driven by host immune barrier. In the prediction of variation, these sites also show a certain role.
3. in order to further characterize the variation of H3 sequences, this study draw the phylogenetic tree. The analysis results showed that the phylogenetic tree structure and clustering results were highly consistent with the corresponding results not only describe the basic rules of the influenza virus mutates in decades, but also a clear answer to the following two key problems: the host plays in pigs the effect of variation of human influenza virus is very small, the main antigen memory effect; compared with the rest of the world, South China antigenic drift in H3 subtype did not show the importance of higher, should not be considered the birthplace of the new subtype variants of the crowd.
4. based on the phylogenetic tree structure has been obtained, this study further analyzes the evolutionary tree trunk of positively selected sites regularly, positive selection of Yang model was used for analysis. The results show that the selective pressure of human influenza virus H3 antigen bearing is gradually increasing, and reached the peak in 1995. In 28 positive sites to be selected, most of them are only a section of the evolution period of influenza virus in under positive selection pressure.
Based on the above analysis results, we believe that the influenza virus subtype H3 showed a gradually accelerating trend in the variation of human, which is mainly due to the screening effect caused by the increasingly powerful immune barrier. At present, the core task is to protect the immune strategy of high-risk groups, but this strategy can not stop, but may be accelerated by the emergence of new the mutants. By farther bioinformatics analysis, and in the world to adopt a unified initiative

【學(xué)位授予單位】：復(fù)旦大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2005
【分類(lèi)號(hào)】：R373

【引證文獻(xiàn)】

相關(guān)博士學(xué)位論文 前1條

1 任曉衛(wèi);上海地區(qū)人群甲型流感HA抗原進(jìn)化與基因進(jìn)化關(guān)系研究及H1N1流感潛在免疫顯性位點(diǎn)的篩選[D];復(fù)旦大學(xué);2010年

，

本文編號(hào)：1362584