通常使用的过滤条件:
1、序列为productive
2、核酸序列为3的倍数,氨基酸序列大于4
其他的过滤条件
例如:expression abundance(TPM):
已看到几篇文章在使用,alpha chain TPM < 10 or beta chain TPM < 15, 但不确定是公认标准还是个人经验设置
结果部分:In total, we detected full TCR sequences for 94% (3,792/4,032) T cells, with at least one paired productive TCR a-b chain for subsequent analyses (Table S5). While most cells expressed unique TCR a and b alleles, nonunique a and/or b could be detected in a fraction of T cells. After
eliminating non-productive alleles (e.g., out-of-frame transcripts) or low-abundance TCRs (Figure S6A), we found that 84% (3,174/3,792) contained unique and productive a chains and 94% (3,559/3,792) unique and productive b chains (Figure S6B), in agreement with previous reports
方法部分:
TCR analysis
The TCR sequences for each single T cell were assembled by the TraCeR method from single cell RNA-Seq data, leading to the identification of the CDR3 sequence, the rearranged TCR genes, and their expression abundance (transcripts per million, TPM). First, we discard those cells with no obvious TCR forms. Then we arrange TCR alpha and beta chain respectively with the following steps. The first TCR alpha (beta) chain was defined as follows: 1) keep all single T cells in which only one productive TCR alpha and beta chain was present. 2) if more than one TCR alpha or beta chain were identified in one T cell, we kept only the cells in which a dominant form of alpha and beta was detected. Often, one alpha/beta chain was productive and the other chain was non-productive, or the expression level of one was far higher than the alternative allele, and the productive or dominant form was identified. Next, we filtered out the second TCR alpha chains with TPM less than 10 and beta chains with TPM less than 15 to eliminate the biological and bioinformatics error based on the histogram analysis for the expression distribution (Figure S6A). From a total 4032 cells with successfully assembled TCR sequences, we identified the TCR alpha/beta pairs for 3792 cells.
cite: Zheng C, Zheng L, Yoo JK, et al. Landscape of Infiltrating T Cells in Liver Cancer Revealed by Single-Cell Sequencing. Cell. 2017;169(7):1342-1356.e16. doi:10.1016/j.cell.2017.05.035
TCR analysis
To reduce false positive assembly, we filtered out TCR assemblies with alpha chain TPM < 10 or beta chain TPM < 15…Only productive (that is, in frame) TCR alpha–beta pairs were considered to define the dominant TCR of a single cell.
cite: Guo, Xinyi et al. “Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing.” Nature medicine vol. 24,7 (2018): 978-985. doi:10.1038/s41591-018-0045-3
本文仅仅简单记录了一下首次处理免疫组库数据的一些收获,还待后续补充。
文章目录前言一、免疫组库二、使用步骤1.引入库2.读入数据总结前言刚接触免疫组库数据提示:以下是本篇文章正文内容,下面案例可供参考一、免疫组库We need a minimum of 30x coverage in order to confidently identify unique VDJ sequences as truly unique. In addition, we need to ensure that this coverage amount can be met for
免疫组库数据分析(三):免疫组库数据可视化
在系列文章第二篇《免疫组库数据分析(二):Excel 分析免疫组库数据》中,分析了免疫组库中V基因、J基因、V-J组合的使用频率。在氨基酸水平,分析了CDR3 的氨基酸的长度分布以及20种氨基酸的使用频率;在免疫组库多样性方面,分析了4种不同的多样性指数。
本篇将利用作图软件Graphpad prism 8以及Excel 将上述分析的数据进行可视化,此外利用在线工具分析CDR3 氨基酸保守性,或者两组样本CDR3长度的氨基酸差异。
数据可视化
1. 免疫组库
import matplotlib.pyplot as plt
df = pd.read_csv("/mnt/g/20220309-scBCR/HY01-1F11_ALL.csv",sep=",",low_memory=False)
免疫组库数据分析(一):windows 系统下MiXCR的安装和使用
免疫系统的T细胞或者B细胞免疫组库的多样性主要取决于抗原决定簇CDR3区域的多样性,CDR3区域有部分V基因的3端到J基因的5‘端序列构成,其中包含D基因。因此如何多维度的分析CDR3至关重要。
本系列文章分析小鼠 5’RACE实验数据,并在Windows 系统下用MIXCR进行初步分析,利用Excel进行进一步分析,利用Graphpad prism 8 以及在线绘图网站进行一系列的可视化分析。希望以少代码
具有可选的CDR3重建步骤,该步骤允许从几个不相交的读取中恢复完整的高变区。 同时使用一流的效率来保护免受误报汇编的复杂算法。
组装克隆型,应用几种错误校正算法以消除由PCR和测序错误引起的人为多样性
克隆型可以基于CDR3序列(默认)以及任何其他区域进行组装,包括全长可变序列(从FR1的开始到FR4的结束)
组装完整的TCR / Ig受
免疫算法与遗传算法其实非常相似,但其独特的地方在于,免疫算法用激励度而非亲和度来衡量结果的好坏,而激励度又与抗体密度有关,这就使得密度大的抗体激励度反而小,让免疫算法有全局搜索的能力,不容易陷入局部最优,接下来我就结合代码来讲解。
2.开发环境
【Anaconda + jupyter notebook python 3.7.9】
3.具体实例
现有一函数,定义域为,函数图像
首先导入需要的包
numpy
pandas(这个好像全程没用到)
matplotl...
MIXCR
羊驼(好像是已经免疫过后的)外周血转录组/基因组经多重PCR扩增后,形成特定库并将这些序列重组于表达载体转入噬菌体(噬菌体展示技术),经固相/液相淘选后得到高亲和力的VHH序列库。该序列库再次放大构成高通量测序库,采用PE300测序策略。
paired reads 组装成productive contig
注释contig得到FWR1/CDR1/FWR2/CDR2/FWR3/CDR3/FWR4等信息
得到clonotype
unique protein的统计信息,unqiu
Read subsampling
Reduce the reads for a given barcode to at most 80,000, because more reads don’t help.
Read trimming
Trim off read bases after enrichment primers.
Graph formation
Build a De Bruijn graph using
单细胞实战(三) Cell Ranger使用初探
把sra文件转化成fastq格式,并对fastq格式的文件进行质控。
find ./10X -name "*R1*.gz">id-1.txt
find ./10X -name "*R1*.gz">id-2.txt
cat id-1.txt id-2.txt >id-all.txt
cat id-all.txt| xargs fastqc -t 20 -o ./
得到质控结果。
基本上可以用于下游分析。
接下来就是用 cell
MVC(Model-View-Controller)是一种软件架构模式,用于将应用程序的逻辑分离成三个不同的组件,即模型(Model)、视图(View)和控制器(Controller)。在JavaWeb项目中,可以使用MVC模式对管理员登录进行RSA加密验证,具体步骤如下:
1. Model层:定义管理员实体类和数据访问层接口。管理员实体类包含管理员的账号和密码属性,数据访问层接口定义管理员登录验证方法。
public class Admin {
private String account;
private String password;
// getter和setter方法
public interface AdminDAO {
public boolean validate(String account, String password);
2. View层:定义登录页面,包括账号和密码输入框和登录按钮。在登录页面的JavaScript代码中,使用RSA算法对密码进行加密,并将加密后的密码传递给Controller层。
<form id="login-form" action="#" method="post">
<input type="text" id="account" name="account" placeholder="请输入账号">
<input type="password" id="password" name="password" placeholder="请输入密码">
<button type="button" id="login-btn">登录</button>
</form>
<script src="js/rsa.js"></script>
<script>
var pubkey = "MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQCWx9X7vUVF+JlRgEj8Iz0L7VdJ\n" +
"1yU6jD9+2gGJ/y+U6V5SbJz/2Q5c5G5t5LJ4Af4hXNlNjuyy+4dD8/BZwDmHv/TY\n" +
"f8tG8mlHJF3c+3fQmQGw5M2QOA+5K5J5L5h4x4pB4oXJ0Kj1f5JZ+8t1aCzwhhO\n" +
"y+E+8lF9f3Gq/H2jJQIDAQAB";
var encrypt = new JSEncrypt();
encrypt.setPublicKey(pubkey);
$('#login-btn').click(function() {
var account = $('#account').val().trim();
var password = $('#password').val().trim();
var encryptedPwd = encrypt.encrypt(password);
$.ajax({
type: 'POST',
url: 'login',
data: { account: account, password: encryptedPwd },
success: function(data) {
// 处理登录结果
error: function() {
alert('登录失败');
</script>
3. Controller层:处理登录请求,调用Model层的数据访问层接口进行管理员登录验证。在登录验证方法中,使用RSA算法对加密后的密码进行解密,并与数据库中保存的密码进行比对。
public class AdminController {
private AdminDAO adminDAO = new AdminDAOImpl();
public void login(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
String account = request.getParameter("account");
String encryptedPwd = request.getParameter("password");
String password = decryptRSA(encryptedPwd); // RSA解密
boolean result = adminDAO.validate(account, password);
if (result) {
request.getSession().setAttribute("admin", account);
response.sendRedirect("admin.jsp");
} else {
response.getWriter().write("登录失败");
private String decryptRSA(String encryptedPwd) {
// 使用私钥解密
return "";
以上就是使用MVC模式对管理员登录进行RSA加密验证的主要步骤。需要注意的是,RSA算法需要在前端页面和后端服务器均实现,才能进行加密和解密。为了提高安全性,可以使用HTTPS协议对数据传输进行加密。
机器学习阴性集的选择 —— drug-target interactions (DTIs)
已经变秃何时变强:
机器学习阴性集的选择 —— drug-target interactions (DTIs)
Kivsen:
机器学习阴性集的选择 —— drug-target interactions (DTIs)
不正经的kimol君: