肿瘤驱动基因的分析预测及其在肝癌中功能的初步研究
发布时间：2018-08-06 10:38
【摘要】：研究背景鉴定肿瘤驱动基因(cancer driver genes)一直都是肿瘤学研究的热点,目前已经有多种基于肿瘤基因组学的工具被开发,其中dJ/dS方法从不同角度鉴定了新的肿瘤驱动基因。dJ/dS仅考虑了基因外显子(exon)和内含子(intron)的剪接位点(junction site,J)突变情况,若J位点发生突变,则pre-mRNA剪接过程失败,进而无法产生正常的成熟mRNA,从而造成基因功能丧失(loss of function),最后诱导细胞癌变。基于这个思路,该方法借鉴了遗传学工具dN/dS的计算原理,用 Junction mutation(J)和 Synonymous mutation(S)观察值之比(obs_JS)除以期望值之比(exp_JS),即dJ/dS=obs_JS/exp_JS,若dJ/dS1,则认为有正选择效应;若dJ/dS1则认为有纯化选择作用;如果dJ/dS=1则认为无选择压力。我们认为肿瘤发生过程中驱动基因受到了正选择作用。相比其他方法,dJ/dS更好地控制了背景突变率(background mutation rate,BMR),提高了方法的灵敏度。但仍存在两个不足之处:一、dJ/dS在考虑J突变时只计算了发生在splice donor(GT)和splice acceptor(AG)四个位点上的突变,而有研究表明这四个位点临近的区域发生突变也会引起剪接失败,导致疾病发生,因此将剪接位点周围的区域纳入J的计算会更加准确;二、dJ/dS在计算突变谱(mutation spectrum)时仅考虑了 12种突变类型,但有文献报道紧邻突变位点的碱基N(ATCG)可能极大地影响突变,所以突变谱采用96种类型计算更科学。因此我们在本研究中对dJ/dS的上述两点不足进行改进,开发dJ/dS2.0版本,并结合TCGA(The Cancer Genome Atlas)33种实体瘤的数据对肿瘤驱动基因重新分析预测。驱动基因在肿瘤中的功能研究也是一项非常重要的课题。癌变过程十分复杂,癌基因和抑癌基因在癌变过程中起重要的调控作用,多种因素可以影响癌基因和抑癌基因的突变,有研究报道人种和性别是影响基因突变的重要因素,而且部分基因突变会使得肿瘤预后更差,我们将以肝癌为研究模型来探讨驱动基因在肝癌中的作用。肝癌是世界上第六位常见的恶性肿瘤,死亡率高居第三位,每年有超过70万新发病例,且呈逐年增加趋势。手术切除是目前早期肝癌最有效的治疗手段,但术后五年复发率超过50%。近年随着测序技术的快速发展,肝癌基因组学研究取得了很大进步,为筛选临床诊疗及预后的基因标志物提供了新的技术手段。本研究拟从两个方面开展:①改进dJ/dS方法并用于鉴定肿瘤驱动基因;②以肝癌为研究模型,结合dJ/dS2.0、dT/dS和MutSig2.0的预测结果,初步分析驱动基因在肝癌中的功能。研究目的1.改进dJ/dS算法,开发dJ/dS2.0版本,并将之用于dJ/dS2.0肿瘤驱动基因的预测;2.初步研究驱动基因在肝癌中的功能。研究方法(一)dJ/dS算法的改进1.dJ/dS2.所用数据:本次研究纳入TCGA的33种实体瘤的数据。2.dJ/dS2.0的计算方法a)计算突变谱:以不受选择压力影响的四倍简并位点上的突变计算每种肿瘤的突变谱,同时采用96种突变类型表示突变谱,即考虑突变位点前后的碱基,如 NCNNTN,N 代表 ATCG;b)J突变的计算原则:在原dJ/dS方法J的基础上还纳入了紧邻剪接位点的外显子上3个碱基和内含子上6个碱基,共11个突变位点;c)S突变的计算原则:纳入每个基因上所有同义突变;d)计算期望JS比值(exp_JS):采用上一步得到的突变谱分别计算每个基因的J和S的期望值(expectation,exp),然后用J的期望值比上S的期望值,即exp_JS = exp_J/exp_S;e)计算实际观察JS比值(obs_JS):从TCGA外显子测序数据中分别统计J和S的观察值(observation,obs),然后用J的观察值比上S的观察值,即obs_JS=obs_J/obs_S;f)计算dJ/dS比值:用实际观察的JS比值除以期望的JS比值得到dJ/dS比值,即dJ/dS obs_J S/exp_J S;(二)驱动基因在肝癌中的功能研究1.肝癌数据来源:TCGA和ICGC;2.驱动基因和通路的鉴定:综合dJ/dS2.0,dT/dS和MutSig2.0三种方法在肝癌数据中的鉴定结果,并利用得到的驱动基因进行通路分析(pathway analysis);3.驱动基因的功能分析:结合肝癌临床资料,分析驱动基因和通路在人种和性别间的突变差异,以及对肝癌预后的影响。(三)统计学分析1.dJ/dS2.0统计分析a)p和FDR计算:使用二项分布检验计算出每个基因对应的p值,同时用Benjamini-Hochberg 方法控制 FDR(false discovery rate);当基因的dJ/dS1且FDR小于0.05时,认为是驱动基因;b)GO功能分析:使用GOrilla进行GO功能聚类分析,FDR小于0.05认为有统计学意义。2.驱动基因功能的统计分析a)KEGG功能分析:利用DAVID运行KEGG功能分析,FDR0.1认为有统计学意义;b)使用fisher确切概率法比较驱动基因及通路在人种和性别间的分布;p0.05认为有统计学意义;c)生存分析:通过Kaplan-Meier生存曲线比较驱动基因对预后的影响,使用log-rank检验计算p值;用Cox比例风险回归模型进行多因素分析;p0.05认为有统计学意义。研究结果(一)dJ/dS2.01.33种肿瘤的突变谱通过在剪接区域增加分析位点和拆分突变谱的方式来改进dJ/dS后,开发dJ/dS2.0版本,将之应用于TCGA33种肿瘤中可以得到更为精准的突变谱。a)高频突变类型CT是大多肿瘤的主要突变类型,其中CpGTpG的突变率最高,但在皮肤黑色素瘤中突变率最高的是CpCCpT和TpCTpT。b)肾癌亚型的突变特点肾嫌色细胞癌突变类型主要以CpGTpG为主,肾透明细胞癌以CCGCAG、CCGCTG和GCGGTG为主,而肾乳头状细胞癌在肾透明细胞癌突变类型的基础上,还包括了 CCACAA和CCCCAC。2.驱动基因除了肾上腺皮质癌、胆管癌、FPPP、肾嫌色细胞癌、弥漫性大B细胞淋巴瘤、子宫癌肉瘤和睾丸生殖细胞瘤这7种肿瘤没有鉴定出结果,dJ/dS2.0在26种鉴定出驱动基因,值得注意的是子宫内膜癌出现多达344个基因,我们怀疑TCGA数据可能有异常或部分基因在子宫内膜癌中更容易发生剪接区域突变,造成dJ/dS2.0不敏感而导致过多假阳性,因此我们移除了子宫内膜癌的鉴定结果,只用余下25种肿瘤的结果进行后续的分析。a)25种肿瘤的驱动基因dJ/dS2.0在25种实体瘤中总共鉴定出643个非冗余的肿瘤驱动基因,除了73个被CGC注释为驱动基因外,另有570个(88.8%)是新预测出来的驱动基因;b)GO功能分析将570个驱动基因进行GO聚类分析,结果显示有很多基因聚集在与发展和维持多细胞性相关的 GO term 上(GO:0044243,multicellular organismal catabolic process,p=2.53E-12,FDR=1.52E-8)。(二)驱动基因在肝癌中的功能分析1.肝癌的驱动基因为了更全面地分析驱动基因的功能,我们结合了 TCGA和ICGC中肝癌的数据来增加样本量,在应用dJ/dS2.0的同时,也使用了分别由我们实验室课题组开发的dT/dS和TCGA开发的MutSig2.0方法,三种方法共鉴定出89个驱动基因,经KEGG pathway分析发现这89个基因富集于10条与病毒感染或肿瘤相关的信号通路上;2.肝癌驱动基因突变类型的特点对比发现,肝癌各驱动基因的突变类型构成比差异较大,在突变频率较高的CTNNB1、TP53和RB1基因中,非同义突变所占比超过95%,其中CTNNB1和TP53以错义突变为主,而插入缺失、错义突变、剪接区域突变及无义突变共存于RB1中;3.人种和性别影响驱动基因和通路突变的分布基因TP53和RB1及通路hsa05161和hsa05203在亚裔中更容易突变,男性中的CTNNB1、ALB、PIK3CA、BAP1基因和hsa05200通路发生率高于女性;4.生存分析经单因素筛选和多因素分析,基因KCNB2(p=0.025),KCNJ12(p=0.015),RB1(p=0.038)和 TP53(p=0.01)及乙肝感染通路 hsa05161(p=0.006)若发生突变则患者中位生存时间缩短,死亡风险比增加。结论第一章:dJ/dS2.01.改进dJ/dS算法,提高了准确性;2.由dJ/dS2.0鉴定的驱动基因参与多细胞性的发展和维持环节;第二章:驱动基因在肝癌中的功能分析1.肝癌患者的性别和人种会影响驱动基因突变的分布;2.驱动基因KCNB2、KCNJ12、RB1、TP53和乙肝感染通路hsa05161突变是肝癌预后的不良因素,可能是临床上潜在的基因标志物。
[Abstract]:Cancer driver genes (tumor driven gene genes) has always been a hot spot in oncology research. At present, a variety of tools based on tumor genomics have been developed, in which the dJ/dS method identified the new tumor driven gene.DJ/dS from the splice site of the exon (exon) and Chi Ko (intron) from different angles (J). Unction site, J) mutation, if the J site mutation, then the pre-mRNA splicing process failed, and then can not produce normal mature mRNA, resulting in the loss of gene function (loss of function), and finally induced cell cancerization. Based on this idea, the method uses the principle of the calculation of dN/dS of the legacy tool dN/dS, and uses Junction mutation and Junction. The ratio of ymous mutation (S) observation value (obs_JS) is divided by the ratio of expected values (exp_JS), that is, dJ/dS=obs_JS/exp_JS, if dJ/dS1, there is a positive selection effect; if dJ/dS1 believes that there is a purification selection effect; if dJ/dS=1 believes that there is no selective pressure. We think that the driving gene is positive in the carcinogenesis. Compared to other methods, D, D. J/dS better controls the background mutation rate (background mutation rate, BMR) and improves the sensitivity of the method. However, there are still two shortcomings: first, dJ/dS only calculates the mutation on the splice donor (GT) and splice acceptor (rate) when the J mutation is considered, and there is a study showing that the four loci are near the mutation. It also causes splicing failure, causing disease to occur, so it is more accurate to incorporate the area around the splice site into J; two, dJ/dS only considers 12 mutation types when calculating the mutation spectrum (mutation spectrum), but the base N (ATCG) that is reported to be adjacent to the mutation site may greatly affect the mutation, so 96 kinds of mutation spectrum are used. Type calculation is more scientific. Therefore, in this study, we improve the above two points of dJ/dS, develop the dJ/dS2.0 version, and combine the data of 33 solid tumors of the TCGA (The Cancer Genome Atlas) to reanalyze the tumor driving genes. It is very complicated that oncogene and tumor suppressor gene play an important regulatory role in the process of carcinogenesis. Many factors can affect the mutation of oncogene and tumor suppressor gene. There is a research report that human species and sex are important factors affecting gene mutation, and some gene mutations will make the tumor be worse after the tumor. We will discuss the model of liver cancer as a study model. Liver cancer is the sixth most common malignant tumor in the world, with a high mortality rate of third, more than 70 million new cases each year, and increasing year by year. Surgical resection is the most effective treatment for early liver cancer, but the recurrence rate of five years after surgery is more than 50%. in recent years, with the rapid development of sequencing technology. Great progress has been made in the study of hepatoma genomics, which provides new technical means for screening gene markers for clinical diagnosis and prognosis. This study is intended to be carried out in two aspects: (1) improving the dJ/dS method and identifying the tumor driven genes; (2) hepatocellular carcinoma as a research model, combined with the prediction results of dJ/ dS2.0, dT/dS and MutSig2.0, preliminary analysis The function of driving genes in liver cancer. Purpose 1. to improve the dJ/dS algorithm, to develop the dJ/dS2.0 version, and to apply it to the prediction of dJ/dS2.0 tumor driven genes; 2. to study the function of the driving genes in the liver cancer. Research method (1) the data of the improved 1.dJ/dS2. for the dJ/dS algorithm: This study included the data.2.dJ/ of 33 kinds of solid tumors of TCGA. DS2.0 calculation method a) calculation of mutation spectrum: the mutation spectrum of each tumor is calculated at four times of degenerate loci without the influence of selected pressure. At the same time, 96 mutation types are used to express the mutation spectrum, that is, the base of the mutation site, such as NCNNTN, N representing ATCG; b) J process, is calculated on the basis of the original dJ/dS method J 3 bases and 6 bases on the exons adjacent to the splice site, a total of 11 mutation sites, and a total of 11 mutation sites; c) S mutation calculation principle: integrating all the synonymous mutations on each gene; d) calculating the expected JS ratio (exp_JS): the expectation of each gene's J and S (expectation, exp) by the previous mutation spectrum, and then J, respectively, are used for J. The expectation value is compared with the expected value of the upper S, that is, exp_JS = exp_J/exp_S; E) to calculate the actual JS ratio (obs_JS): the observed values of J and S (observation, OBS) from the TCGA exons sequencing data are calculated, and then the ratio of the actual observed values is divided by the observed values of J. The ratio obtained dJ/dS ratio, dJ/dS obs_J S/exp_J S; (two) the function of the driving genes in the liver cancer 1. the source of liver cancer data: TCGA and ICGC; the identification of 2. driving genes and pathways: comprehensive dJ/dS2.0, dT/dS and MutSig2.0 three methods in the identification of liver cancer data, and using the obtained driving genes for pathway analysis (pathway analy) SIS): functional analysis of 3. driving genes: combined with the clinical data of liver cancer, analysis of the mutation difference between the driving genes and pathways in human and sex, and the effect on the prognosis of liver cancer. (three) statistical analysis of 1.dJ/dS2.0 statistical analysis a) P and FDR calculation: using two distribution tests to calculate the corresponding p value of each gene, and use Benjamini-Hochberg at the same time. The method controls FDR (false discovery rate); when the gene is dJ/dS1 and FDR is less than 0.05, it is considered to be the driving gene; b) GO functional analysis: GO function clustering analysis using GOrilla, FDR less than 0.05 Statistical significance; b) using the Fisher exact probability method to compare the distribution of driving genes and pathways between human and sex; P0.05 considered statistically significant; c) survival analysis: using the Kaplan-Meier survival curve to drive the gene to the prognosis of the gene, using the log-rank test to calculate the p value, and using the Cox proportional risk regression model for multifactor analysis; P 0.05 (0.05) the results (1) the mutation spectrum of the dJ/dS2.01.33 tumor is improved by increasing the analysis site and splitting the mutation spectrum in the splicing region, developing the dJ/dS2.0 version and applying it to the TCGA33 tumor to get a more accurate mutation spectrum.A) the high frequency mutation type CT is the majority of the tumors. In the case of mutation, the mutation rate of CpGTpG is the highest, but the highest mutation rate in the skin melanoma is CpCCpT and TpCTpT.b). The mutation of renal carcinoma subtypes is characterized mainly by CpGTpG, and the renal clear cell carcinoma is mainly CCGCAG, CCGCTG and GCGGTG, while the renal papillary cell carcinoma is in the renal clear cell carcinoma mutation type. On the basis of the CCACAA and CCCCAC.2. driving genes, 7 kinds of tumors, including adrenocortical cancer, cholangiocarcinoma, FPPP, renal chromophobe cell carcinoma, diffuse large B cell lymphoma, uterine carcinosarcoma and testicular germ cell tumor, were not identified, and dJ/dS2.0 was identified in 26 of the driving genes. It is worth noting that endometrial cancer appears much more. Up to 344 genes, we suspect that TCGA data may have abnormal or part of the gene in endometrial cancer more prone to splicing region mutation, resulting in dJ/dS2.0 insensitivity and lead to excessive false positive. Therefore, we remove the identification of endometrial cancer and use only the results of the remaining 25 tumors for subsequent analysis of.A) the driving of 25 kinds of tumors. The gene dJ/dS2.0 identified 643 non redundant tumor driven genes in 25 solid tumors. In addition to 73 CGC annotated as driving genes, another 570 (88.8%) was a newly predicted drive gene; b) GO functional analysis carried out GO clustering analysis of 570 driving genes. The results showed that many genes were aggregated in development and maintenance. Cytoplasmic related GO term (GO:0044243, multicellular organismal catabolic process, p=2.53E-12, FDR=1.52E-8). (two) functional analysis of the driving genes in liver cancer 1. the driving genes of the liver cancer in order to more comprehensively analyze the function of the driving genes, we combine the data of the liver cancer in TCGA and ICGC to increase the sample size, in the application of dJ/dS2. 0 at the same time, we also used the MutSig2.0 method developed by our lab group dT/dS and TCGA respectively. Three methods were used to identify 89 driving genes. The 89 genes were found to be enriched in 10 signal pathways related to virus infection or tumor by KEGG pathway analysis. The characteristics of the mutation types of the 2. liver cancer driving genes were compared. At present, the mutation types of liver cancer drive genes are different in proportion. In CTNNB1, TP53 and RB1 genes with higher mutation frequency, the proportion of non synonymous mutations is more than 95%, of which CTNNB1 and TP53 are mainly missense mutations, while insertion loss, missense mutation, splicing region mutation and nonsense mutation coexist in RB1; 3. and sex influence drivers Distribution genes TP53 and RB1 and pathway hsa05161 and hsa05203 are more likely to be mutated in Asian descent. The incidence of CTNNB1, ALB, PIK3CA, BAP1 and hsa05200 pathways in men is higher than that of women; 4. survival analysis, by single factor screening and multifactor analysis, gene KCNB2 (p=0.025), KCNJ12 (KCNJ12), etc., and If the hepatitis B infection pathway hsa05161 (p=0.006) mutation occurs, the median survival time of the patients is shortened and the mortality risk is increased. Conclusion the first chapter: dJ/dS2.01. improves the dJ/dS algorithm and improves the accuracy; 2. the driving genes identified by dJ/dS2.0 are involved in the development of multicellular and maintenance rings; and the second chapter: functional analysis of the driving gene in liver cancer 1. The sex and ethnicity of the patients with liver cancer may affect the distribution of the driving gene mutation; the 2. driven gene KCNB2, KCNJ12, RB1, TP53 and the hsa05161 mutation in the hepatitis B infection pathway are adverse factors for the prognosis of liver cancer, which may be a potentially clinical marker.
【学位授予单位】：南方医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R735.7
本文编号：2167473
[Abstract]:Cancer driver genes (tumor driven gene genes) has always been a hot spot in oncology research. At present, a variety of tools based on tumor genomics have been developed, in which the dJ/dS method identified the new tumor driven gene.DJ/dS from the splice site of the exon (exon) and Chi Ko (intron) from different angles (J). Unction site, J) mutation, if the J site mutation, then the pre-mRNA splicing process failed, and then can not produce normal mature mRNA, resulting in the loss of gene function (loss of function), and finally induced cell cancerization. Based on this idea, the method uses the principle of the calculation of dN/dS of the legacy tool dN/dS, and uses Junction mutation and Junction. The ratio of ymous mutation (S) observation value (obs_JS) is divided by the ratio of expected values (exp_JS), that is, dJ/dS=obs_JS/exp_JS, if dJ/dS1, there is a positive selection effect; if dJ/dS1 believes that there is a purification selection effect; if dJ/dS=1 believes that there is no selective pressure. We think that the driving gene is positive in the carcinogenesis. Compared to other methods, D, D. J/dS better controls the background mutation rate (background mutation rate, BMR) and improves the sensitivity of the method. However, there are still two shortcomings: first, dJ/dS only calculates the mutation on the splice donor (GT) and splice acceptor (rate) when the J mutation is considered, and there is a study showing that the four loci are near the mutation. It also causes splicing failure, causing disease to occur, so it is more accurate to incorporate the area around the splice site into J; two, dJ/dS only considers 12 mutation types when calculating the mutation spectrum (mutation spectrum), but the base N (ATCG) that is reported to be adjacent to the mutation site may greatly affect the mutation, so 96 kinds of mutation spectrum are used. Type calculation is more scientific. Therefore, in this study, we improve the above two points of dJ/dS, develop the dJ/dS2.0 version, and combine the data of 33 solid tumors of the TCGA (The Cancer Genome Atlas) to reanalyze the tumor driving genes. It is very complicated that oncogene and tumor suppressor gene play an important regulatory role in the process of carcinogenesis. Many factors can affect the mutation of oncogene and tumor suppressor gene. There is a research report that human species and sex are important factors affecting gene mutation, and some gene mutations will make the tumor be worse after the tumor. We will discuss the model of liver cancer as a study model. Liver cancer is the sixth most common malignant tumor in the world, with a high mortality rate of third, more than 70 million new cases each year, and increasing year by year. Surgical resection is the most effective treatment for early liver cancer, but the recurrence rate of five years after surgery is more than 50%. in recent years, with the rapid development of sequencing technology. Great progress has been made in the study of hepatoma genomics, which provides new technical means for screening gene markers for clinical diagnosis and prognosis. This study is intended to be carried out in two aspects: (1) improving the dJ/dS method and identifying the tumor driven genes; (2) hepatocellular carcinoma as a research model, combined with the prediction results of dJ/ dS2.0, dT/dS and MutSig2.0, preliminary analysis The function of driving genes in liver cancer. Purpose 1. to improve the dJ/dS algorithm, to develop the dJ/dS2.0 version, and to apply it to the prediction of dJ/dS2.0 tumor driven genes; 2. to study the function of the driving genes in the liver cancer. Research method (1) the data of the improved 1.dJ/dS2. for the dJ/dS algorithm: This study included the data.2.dJ/ of 33 kinds of solid tumors of TCGA. DS2.0 calculation method a) calculation of mutation spectrum: the mutation spectrum of each tumor is calculated at four times of degenerate loci without the influence of selected pressure. At the same time, 96 mutation types are used to express the mutation spectrum, that is, the base of the mutation site, such as NCNNTN, N representing ATCG; b) J process, is calculated on the basis of the original dJ/dS method J 3 bases and 6 bases on the exons adjacent to the splice site, a total of 11 mutation sites, and a total of 11 mutation sites; c) S mutation calculation principle: integrating all the synonymous mutations on each gene; d) calculating the expected JS ratio (exp_JS): the expectation of each gene's J and S (expectation, exp) by the previous mutation spectrum, and then J, respectively, are used for J. The expectation value is compared with the expected value of the upper S, that is, exp_JS = exp_J/exp_S; E) to calculate the actual JS ratio (obs_JS): the observed values of J and S (observation, OBS) from the TCGA exons sequencing data are calculated, and then the ratio of the actual observed values is divided by the observed values of J. The ratio obtained dJ/dS ratio, dJ/dS obs_J S/exp_J S; (two) the function of the driving genes in the liver cancer 1. the source of liver cancer data: TCGA and ICGC; the identification of 2. driving genes and pathways: comprehensive dJ/dS2.0, dT/dS and MutSig2.0 three methods in the identification of liver cancer data, and using the obtained driving genes for pathway analysis (pathway analy) SIS): functional analysis of 3. driving genes: combined with the clinical data of liver cancer, analysis of the mutation difference between the driving genes and pathways in human and sex, and the effect on the prognosis of liver cancer. (three) statistical analysis of 1.dJ/dS2.0 statistical analysis a) P and FDR calculation: using two distribution tests to calculate the corresponding p value of each gene, and use Benjamini-Hochberg at the same time. The method controls FDR (false discovery rate); when the gene is dJ/dS1 and FDR is less than 0.05, it is considered to be the driving gene; b) GO functional analysis: GO function clustering analysis using GOrilla, FDR less than 0.05 Statistical significance; b) using the Fisher exact probability method to compare the distribution of driving genes and pathways between human and sex; P0.05 considered statistically significant; c) survival analysis: using the Kaplan-Meier survival curve to drive the gene to the prognosis of the gene, using the log-rank test to calculate the p value, and using the Cox proportional risk regression model for multifactor analysis; P 0.05 (0.05) the results (1) the mutation spectrum of the dJ/dS2.01.33 tumor is improved by increasing the analysis site and splitting the mutation spectrum in the splicing region, developing the dJ/dS2.0 version and applying it to the TCGA33 tumor to get a more accurate mutation spectrum.A) the high frequency mutation type CT is the majority of the tumors. In the case of mutation, the mutation rate of CpGTpG is the highest, but the highest mutation rate in the skin melanoma is CpCCpT and TpCTpT.b). The mutation of renal carcinoma subtypes is characterized mainly by CpGTpG, and the renal clear cell carcinoma is mainly CCGCAG, CCGCTG and GCGGTG, while the renal papillary cell carcinoma is in the renal clear cell carcinoma mutation type. On the basis of the CCACAA and CCCCAC.2. driving genes, 7 kinds of tumors, including adrenocortical cancer, cholangiocarcinoma, FPPP, renal chromophobe cell carcinoma, diffuse large B cell lymphoma, uterine carcinosarcoma and testicular germ cell tumor, were not identified, and dJ/dS2.0 was identified in 26 of the driving genes. It is worth noting that endometrial cancer appears much more. Up to 344 genes, we suspect that TCGA data may have abnormal or part of the gene in endometrial cancer more prone to splicing region mutation, resulting in dJ/dS2.0 insensitivity and lead to excessive false positive. Therefore, we remove the identification of endometrial cancer and use only the results of the remaining 25 tumors for subsequent analysis of.A) the driving of 25 kinds of tumors. The gene dJ/dS2.0 identified 643 non redundant tumor driven genes in 25 solid tumors. In addition to 73 CGC annotated as driving genes, another 570 (88.8%) was a newly predicted drive gene; b) GO functional analysis carried out GO clustering analysis of 570 driving genes. The results showed that many genes were aggregated in development and maintenance. Cytoplasmic related GO term (GO:0044243, multicellular organismal catabolic process, p=2.53E-12, FDR=1.52E-8). (two) functional analysis of the driving genes in liver cancer 1. the driving genes of the liver cancer in order to more comprehensively analyze the function of the driving genes, we combine the data of the liver cancer in TCGA and ICGC to increase the sample size, in the application of dJ/dS2. 0 at the same time, we also used the MutSig2.0 method developed by our lab group dT/dS and TCGA respectively. Three methods were used to identify 89 driving genes. The 89 genes were found to be enriched in 10 signal pathways related to virus infection or tumor by KEGG pathway analysis. The characteristics of the mutation types of the 2. liver cancer driving genes were compared. At present, the mutation types of liver cancer drive genes are different in proportion. In CTNNB1, TP53 and RB1 genes with higher mutation frequency, the proportion of non synonymous mutations is more than 95%, of which CTNNB1 and TP53 are mainly missense mutations, while insertion loss, missense mutation, splicing region mutation and nonsense mutation coexist in RB1; 3. and sex influence drivers Distribution genes TP53 and RB1 and pathway hsa05161 and hsa05203 are more likely to be mutated in Asian descent. The incidence of CTNNB1, ALB, PIK3CA, BAP1 and hsa05200 pathways in men is higher than that of women; 4. survival analysis, by single factor screening and multifactor analysis, gene KCNB2 (p=0.025), KCNJ12 (KCNJ12), etc., and If the hepatitis B infection pathway hsa05161 (p=0.006) mutation occurs, the median survival time of the patients is shortened and the mortality risk is increased. Conclusion the first chapter: dJ/dS2.01. improves the dJ/dS algorithm and improves the accuracy; 2. the driving genes identified by dJ/dS2.0 are involved in the development of multicellular and maintenance rings; and the second chapter: functional analysis of the driving gene in liver cancer 1. The sex and ethnicity of the patients with liver cancer may affect the distribution of the driving gene mutation; the 2. driven gene KCNB2, KCNJ12, RB1, TP53 and the hsa05161 mutation in the hepatitis B infection pathway are adverse factors for the prognosis of liver cancer, which may be a potentially clinical marker.
【学位授予单位】：南方医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R735.7
【相似文献】
相关期刊论文 前10条
1 欧阳军;;基因与长寿[J];家庭中医药;2006年10期
2 文辉;人到底有多少个基因?[J];百科知识;2003年11期
3 张丽华;侯振江;;p53基因在肺癌研究中的进展[J];临床肺科杂志;2006年01期
4 程鹏,童茂荣;p53基因在肺癌的治疗及预后中的价值[J];江苏医药;2000年03期
5 卢薇薇;仇容;;nm23基因的研究进展[J];中国实用医药;2008年30期
6 张爱菊;P~(53)基因与人类肿瘤[J];医学综述;1998年08期
7 熊建新,吴子忠;p53基因与肺癌[J];中国煤炭工业医学杂志;1999年01期
8 姚松朝;李辉;;p53基因与肺癌[J];海军医学;1995年01期
9 王武;卢辉山;;KAI1基因与消化道恶性肿瘤关系的研究现状[J];医学综述;2007年07期
10 马秀华,李翠萍,王建伟;Rb与P~(16)基因在肺癌组织中的表达及其意义[J];山东医药;1999年19期
相关会议论文 前10条
1 张蕊;Liang Zhao;Tom Gonda;;慢病毒介导的多西环素可诱导式短发夹RNA抑制K562细胞系e-myb基因的表达[A];中华医学会第十四次全国儿科学术会议论文汇编[C];2006年
2 高艳红;岳文;秦立鹏;张鹏;陈琳;裴雪涛;;人Spindlin1基因的功能初探[A];第九届全国实验血液学会议论文摘要汇编[C];2003年
3 王卫东;陈正堂;李德志;;肿瘤治疗新策略——基因—放射治疗:原理、现状及展望[A];中华医学会放射肿瘤治疗学分会六届二次暨中国抗癌协会肿瘤放疗专业委员会二届二次学术会议论文集[C];2009年
4 王今达;;21世纪生命科学飞跃发展的基础——基因工程[A];2000年全国危重病急救医学学术会议论文集[C];2000年
5 李宝健;许新萍;冯道荣;范钦;朱华晨;林莉;;论改良生物体遗传性的多基因策略——Ⅰ.多基因改良策略将成为未来基因工程的主流方向[A];中国生物工程学会第三次全国会员代表大会暨学术讨论会论文摘要集[C];2001年
6 王根轩;;基因表达格局与植物数量性状的控制[A];全国植物分子生物学与生物技术学术研讨会论文集[C];2000年
7 彭志宏;黄柏胜;文芳;韩仰;刘发益;邬力祥;;Mettl9基因对U251细胞生长的影响及机制研究[A];湖南省生理科学会2013年度学术年会论文摘要汇编[C];2013年
8 付凌洁;张淑兰;;RASSF1A基因在卵巢癌中的表达及其对增殖、凋亡的影响[A];东北三省第四届妇产科学术会议论文汇编[C];2008年
9 钱伟;何云刚;赖建华;舒坤贤;谭德勇;余敏;;p53基因对mybbp1a(p160)基因的正调控作用研究[A];中国生物化学与分子生物学会第八届会员代表大会暨全国学术会议论文摘要集[C];2001年
10 薛世伟;林艳丽;熊福银;吴晓洁;周艳荣;郭虹敏;陈红星;陈树林;;人Nanog基因的克隆及其在CHO-K1细胞中的表达[A];中国畜牧兽医学会动物解剖学及组织胚胎学分会第十六次学术研讨会论文集[C];2010年
相关重要报纸文章 前10条
1 全国政协委员 民盟中央委员 曹义孙;立法规范基因检测[N];团结报;2010年
2 吕回;鸟儿会唱人会说 只缘基因太相似[N];广东科技报;2004年
3 ;认识我们的基因[N];科技日报;2003年
4 本报记者 张丽;基因可以透视明天吗[N];中国消费者报;2001年
5 姜海;科学家发现新的致盲基因[N];科技日报;2007年
6 徐阳;基因与长寿[N];光明日报;2000年
7 记者 林小春 任海军;美国研究表明懒或许源于基因[N];新华每日电讯;2013年
8 卢菁;怕胖可就是管不住嘴,原是基因作怪[N];新华每日电讯;2008年
9 本报记者 卢雁妮;“从对生命及疾病本源的认识到基因科技造福人类”讲座开讲[N];黔西南日报;2014年
10 张小军;美发现老鼠性别新基因[N];医药经济报;2001年
相关博士学位论文 前10条
1 刘芳;RB94基因在非小细胞肺癌治疗中的作用及机制研究[D];山东大学;2015年
2 王宗丹;远程调控元件mbr选择性协调BCL2和NOXA基因转录活性的机制研究[D];南京医科大学;2013年
3 王晓乐;水稻脆鞘基因BSH1的图位克隆与功能分析[D];中国农业科学院;2015年
4 刘莉莉;奶牛产奶性状候选功能基因VPS28的功能验证[D];中国农业大学;2016年
5 王义涛;ACER2是一个新的通过促ROS产生而诱发自噬与凋亡的p53靶基因[D];重庆医科大学;2017年
6 廖传文;WTX基因调控肝癌细胞增殖及侵袭转移的机制研究[D];南昌大学;2016年
7 金旭;心血管内靶向定位基因递送体系—载基因支架的实验研究[D];中国协和医科大学;2006年
8 杨春霞;生活方式和基因多态与食道癌的关系[D];四川大学;2005年
9 吴汉林;人NIT1基因cDNA的克隆、原核表达及抗体制备[D];四川大学;2005年
10 何勇;p73基因在人肺癌中的表达及对其细胞生物学行为的影响[D];第三军医大学;2004年
相关硕士学位论文 前10条
1 邓小冬;ERCC1、XPF和XPG基因在法医学年龄推断中的初步研究[D];川北医学院;2015年
2 张润贤;论基因权的私权属性及民法保护[D];东北林业大学;2015年
3 刘伟;家蚕受体酪氨酸激酶样的孤独受体Ror2基因功能的研究[D];苏州大学;2015年
4 卫姣;铜离子感应基因MAC1转录机理的研究[D];郑州大学;2015年
5 杨镇宇;禁食—恢复投喂条件下鲢神经肽Y及三个瘦素相关基因的表达分析[D];华中农业大学;2015年
6 周彦宽;Fam49b基因在人类细胞系中的定向敲除及其功能研究[D];浙江大学;2015年
7 刘如锦;病毒巨噬细胞炎症蛋白Ⅱ调控抗HIV-1基因APOBEC3G表达机制的初步探讨[D];杭州师范大学;2015年
8 罗萌萌;HEPIS基因功能的初步研究[D];华北理工大学;2015年
9 梁一丹;应用双荧光素酶报告载体研究上游开放阅读框的突变对靶基因的影响[D];南方医科大学;2015年
10 李霞;miRNA靶基因优选及网络分析[D];天津医科大学;2015年
，本文编号：2167473
本文链接：https://www.wllwen.com/kejilunwen/jiyingongcheng/2167473.html
最近更新
教材专著
