"use strict"; const {Fragment: _Fragment, jsx: _jsx, jsxs: _jsxs} = arguments[0]; const {useMDXComponents: _provideComponents} = arguments[0]; function _createMdxContent(props) { const _components = { a: "a", blockquote: "blockquote", code: "code", h1: "h1", h2: "h2", h3: "h3", img: "img", li: "li", p: "p", path: "path", pre: "pre", span: "span", strong: "strong", svg: "svg", ul: "ul", ..._provideComponents(), ...props.components }; return _jsxs(_Fragment, { children: [_jsxs(_components.p, { children: ["质控 -> 过滤", _jsx(_components.code, { className: "language-auto", children: "subset()" }), " -> 标准化", _jsx(_components.code, { className: "language-auto", children: "NormalizeData()" }), " -> 识别HVGs", _jsx(_components.code, { className: "language-auto", children: "FindVariableFeatures()" }), " -> 标准化", _jsx(_components.code, { className: "language-auto", children: "ScaleData()" }), " -> 聚类", _jsx(_components.code, { className: "language-auto", children: "FindClusters/FindNeighbors" }), " -> 作图", _jsx(_components.code, { className: "language-auto", children: "RunUMAP()" }), " ->"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["rm", _jsx(_components.span, { className: "token punctuation", children: "(" }), "list ", _jsx(_components.span, { className: "token operator", children: "=" }), " ls", _jsx(_components.span, { className: "token punctuation", children: "(" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["pbmc.data ", _jsx(_components.span, { className: "token operator", children: "<-" }), " Read10X", _jsx(_components.span, { className: "token punctuation", children: "(" }), "data.dir ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"filtered_gene_bc_matrices/hg19/\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " suppressWarnings", _jsx(_components.span, { className: "token punctuation", children: "(" }), "CreateSeuratObject", _jsx(_components.span, { className: "token punctuation", children: "(" }), "counts ", _jsx(_components.span, { className: "token operator", children: "=" }), " pbmc.data", _jsx(_components.span, { className: "token punctuation", children: "," }), " project ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"pbmc3k\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " min.cells ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "3" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " min.features ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "200" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["ncol", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "#有多少个细胞，ncol(pbmc.data)亦可 " }), "\n"] })] }) }), "\n", _jsxs(_components.h1, { id: "储存的数据结构", children: [_jsx(_components.a, { className: "heading-link", href: "#储存的数据结构", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "储存的数据结构"] }), "\n", _jsxs(_components.h2, { id: "assays", children: [_jsx(_components.a, { className: "heading-link", href: "#assays", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "Assays"] }), "\n", _jsxs(_components.p, { children: ["储存表达矩阵。默认情况下，我们是对Seurat中的RNA的Assay进行操作。可以通过", _jsx(_components.code, { className: "language-auto", children: "@active.assay" }), "查看当前默认的assay，通过DefaultAssay()更改当前的默认assay。"] }), "\n", _jsx(_components.p, { children: "counts 存储原始数据，是稀疏矩阵" }), "\n", _jsx(_components.p, { children: "data存储logNormalize() 规范化的data。总表达式对每个单元格的要素表达式度量进行标准化，将其乘以比例因子（默认为10,000），并对结果进行对数转换" }), "\n", _jsx(_components.p, { children: "scale.data存储ScaleData()缩放后的data，此步骤需要时间久。" }), "\n", _jsxs(_components.h2, { id: "metadata", children: [_jsx(_components.a, { className: "heading-link", href: "#metadata", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "meta.data"] }), "\n", _jsx(_components.p, { children: "元数据，对每个细胞的描述。一般的meta.data包括orig.ident, nCount_RNA, nFeature_RNA, 以及计算后的percent.mt，RNA_snn_res.0.5等" }), "\n", _jsx(_components.p, { children: "使用时有所区别：" }), "\n", _jsxs(_components.ul, { children: ["\n", _jsxs(_components.li, { children: [_jsx(_components.code, { className: "language-auto", children: "pbmc[['nCount_RNA']]" }), " 取出所有细胞的nCount_RNA，包含细胞信息，数据框"] }), "\n", _jsxs(_components.li, { children: ["而", _jsx(_components.code, { className: "language-auto", children: "pbmc@meta.data$nCount_RNA" }), "取出来的是单独nCount_RNA一列，是向量"] }), "\n"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["head", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token string", children: "'nCount_RNA'" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: " nCount_RNA\n" }), _jsxs(_components.span, { className: "code-line", children: ["AAACATACAACCAC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token number", children: "2419" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACATTGAGCTAC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token number", children: "4903" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACATTGATCAGC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token number", children: "3147" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACCGTGCTTCCG", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token number", children: "2639" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACCGTGTATGCG", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token number", children: "980" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACGCACTGGTAC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token number", children: "2163" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" }), _jsxs(_components.span, { className: "code-line", children: ["head", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token operator", children: "@" }), "meta.data", _jsx(_components.span, { className: "token operator", children: "$" }), "nCount_RNA", _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), " ", _jsx(_components.span, { className: "token number", children: "2419" }), " ", _jsx(_components.span, { className: "token number", children: "4903" }), " ", _jsx(_components.span, { className: "token number", children: "3147" }), " ", _jsx(_components.span, { className: "token number", children: "2639" }), " ", _jsx(_components.span, { className: "token number", children: "980" }), " ", _jsx(_components.span, { className: "token number", children: "2163" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" })] }) }), "\n", _jsxs(_components.h1, { id: "数据质控", children: [_jsx(_components.a, { className: "heading-link", href: "#数据质控", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "数据质控"] }), "\n", _jsx(_components.p, { children: "通常使用的QC指标：" }), "\n", _jsxs(_components.ul, { children: ["\n", _jsx(_components.li, { children: "每个细胞被检测到的unique genes数量（低质量的细胞或空的液滴含有unique genes较少；细胞双重态或多重态检测到异常高unique genes）；" }), "\n", _jsx(_components.li, { children: "细胞内被检测到的features的总数目（与unique genes高度相关）；" }), "\n", _jsx(_components.li, { children: "线粒体基因占比（低质量/将要死去的细胞经常出现线粒体基因占比过高）。" }), "\n"] }), "\n", _jsxs(_components.p, { children: ["其中人类参考基因中线粒体基因是以“MT-”开头的，而在小鼠中是以“mt-”开头的（", _jsx(_components.code, { className: "language-auto", children: "-" }), "是不可少的）\nPs: ", _jsx(_components.code, { className: "language-auto", children: "[]" }), "用于索引向量、列表、数据框。 ", _jsx(_components.code, { className: "language-auto", children: "[[]]" }), "用于索引获得列表、数据框的具体值。"] }), "\n", _jsx(_components.p, { children: "查看线粒体基因占比：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc", _jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token string", children: "\"percent.mt\"" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), " ", _jsx(_components.span, { className: "token operator", children: "<-" }), " PercentageFeatureSet", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " pattern ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"^MT-\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["head", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token operator", children: "@" }), "meta.data", _jsx(_components.span, { className: "token punctuation", children: "," }), _jsx(_components.span, { className: "token number", children: "5" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" }), _jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token operator", children: "-" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: " \n" }), _jsx(_components.span, { className: "code-line", children: " \n" }), _jsx(_components.span, { className: "code-line", children: " orig.ident nCount_RNA nFeature_RNA percent.mt\n" }), _jsxs(_components.span, { className: "code-line", children: ["AAACATACAACCAC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), "\tpbmc3k\t", _jsx(_components.span, { className: "token number", children: "2419" }), "\t", _jsx(_components.span, { className: "token number", children: "779" }), "\t", _jsx(_components.span, { className: "token number", children: "3.0177759" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACATTGAGCTAC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), "\tpbmc3k\t", _jsx(_components.span, { className: "token number", children: "4903" }), "\t", _jsx(_components.span, { className: "token number", children: "1352" }), "\t", _jsx(_components.span, { className: "token number", children: "3.7935958" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACATTGATCAGC", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), "\tpbmc3k\t", _jsx(_components.span, { className: "token number", children: "3147" }), "\t", _jsx(_components.span, { className: "token number", children: "1129" }), "\t", _jsx(_components.span, { className: "token number", children: "0.8897363" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACCGTGCTTCCG", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), "\tpbmc3k\t", _jsx(_components.span, { className: "token number", children: "2639" }), "\t", _jsx(_components.span, { className: "token number", children: "960" }), "\t", _jsx(_components.span, { className: "token number", children: "1.7430845" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["AAACCGTGTATGCG", _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token number", children: "1" }), "\tpbmc3k\t", _jsx(_components.span, { className: "token number", children: "980" }), "\t", _jsx(_components.span, { className: "token number", children: "521" }), "\t", _jsx(_components.span, { className: "token number", children: "1.2244898" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" })] }) }), "\n", _jsxs(_components.h2, { id: "可视化", children: [_jsx(_components.a, { className: "heading-link", href: "#可视化", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "可视化："] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["VlnPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " features ", _jsx(_components.span, { className: "token operator", children: "=" }), " c", _jsx(_components.span, { className: "token punctuation", children: "(" }), _jsx(_components.span, { className: "token string", children: "\"nFeature_RNA\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"nCount_RNA\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"percent.mt\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ncol ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "3" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.1.png", alt: "", className: "internal" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["plot1 ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FeatureScatter", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " feature1 ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"nCount_RNA\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " feature2 ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"percent.mt\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["plot2 ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FeatureScatter", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " feature1 ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"nCount_RNA\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " feature2 ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"nFeature_RNA\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["plot1 ", _jsx(_components.span, { className: "token operator", children: "+" }), " plot2\n"] })] }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.2.png", alt: "", className: "internal" }), "\n", _jsxs(_components.h1, { id: "细胞分群", children: [_jsx(_components.a, { className: "heading-link", href: "#细胞分群", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "细胞分群"] }), "\n", _jsxs(_components.h2, { id: "过滤", children: [_jsx(_components.a, { className: "heading-link", href: "#过滤", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "过滤"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token comment", children: "# 官方推荐过滤掉独特特征计数超过2500或小于200的细胞，或者过滤掉超过5%线粒体基因比例的细胞" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " subset", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " subset ", _jsx(_components.span, { className: "token operator", children: "=" }), " nFeature_RNA ", _jsx(_components.span, { className: "token operator", children: ">" }), " ", _jsx(_components.span, { className: "token number", children: "200" }), " ", _jsx(_components.span, { className: "token operator", children: "&" }), " nFeature_RNA ", _jsx(_components.span, { className: "token operator", children: "<" }), " ", _jsx(_components.span, { className: "token number", children: "2500" }), " ", _jsx(_components.span, { className: "token operator", children: "&" }), " percent.mt ", _jsx(_components.span, { className: "token operator", children: "<" }), " ", _jsx(_components.span, { className: "token number", children: "5" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "pbmc\n" })] }) }), "\n", _jsxs(_components.h2, { id: "标准化", children: [_jsx(_components.a, { className: "heading-link", href: "#标准化", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "标准化"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " NormalizeData", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " normalization.method ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"LogNormalize\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " scale.factor ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "10000" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) }), "\n", _jsxs(_components.p, { children: [_jsx(_components.code, { className: "language-auto", children: "pbmc <- NormalizeData(pbmc)" }), " 亦可，因为上述用的是默认参数。"] }), "\n", _jsx(_components.p, { children: "Ps：如果有了一个TPM的count矩阵，那么就不需要使用Seurat::NormalizeData()操作，因为TPM已经根据测序深度进行了归一化，只需要进行log降维度即可。" }), "\n", _jsxs(_components.h2, { id: "识别差异基因hvgs-highly-variable-features", children: [_jsx(_components.a, { className: "heading-link", href: "#识别差异基因hvgs-highly-variable-features", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "识别差异基因HVGs (highly variable features)"] }), "\n", _jsx(_components.p, { children: "在细胞与细胞间进行比较，选择表达量差别最大的(也即是同一个基因在有的细胞中表达量很高，同时在部分细胞中表达很少)，利用FindVariableFeatures函数，会计算一个mean-variance 结果，也就是给出表达量均值和方差的关系并且得到top variable features。" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FindVariableFeatures", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " selection.method ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"vst\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " nfeatures ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "2000" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# 两千个高变基因" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["top10 ", _jsx(_components.span, { className: "token operator", children: "<-" }), " head", _jsx(_components.span, { className: "token punctuation", children: "(" }), "VariableFeatures", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token number", children: "10" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# 查看十个程度最强的高变基因" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "top10\n" })] }) }), "\n", _jsx(_components.p, { children: "method有三种：" }), "\n", _jsxs(_components.ul, { children: ["\n", _jsx(_components.li, { children: "(默认) vst：对方差(variance)和均值(mean)都取log值+loess线性拟合" }), "\n", _jsx(_components.li, { children: "mean.var.plot方法：average expression & dispersion (for each feature) + z-scores" }), "\n", _jsx(_components.li, { children: "dispersion方法：the highest dispersion values" }), "\n"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["par", _jsx(_components.span, { className: "token punctuation", children: "(" }), "mfrow", _jsx(_components.span, { className: "token operator", children: "=" }), "c", _jsx(_components.span, { className: "token punctuation", children: "(" }), _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["plot1 ", _jsx(_components.span, { className: "token operator", children: "<-" }), " VariableFeaturePlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["plot2 ", _jsx(_components.span, { className: "token operator", children: "<-" }), " LabelPoints", _jsx(_components.span, { className: "token punctuation", children: "(" }), "plot ", _jsx(_components.span, { className: "token operator", children: "=" }), " plot1", _jsx(_components.span, { className: "token punctuation", children: "," }), " points ", _jsx(_components.span, { className: "token operator", children: "=" }), " top10", _jsx(_components.span, { className: "token punctuation", children: "," }), " repel ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token boolean", children: "TRUE" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["combined_plot ", _jsx(_components.span, { className: "token operator", children: "<-" }), " grid.arrange", _jsx(_components.span, { className: "token punctuation", children: "(" }), "plot1", _jsx(_components.span, { className: "token punctuation", children: "," }), " plot2", _jsx(_components.span, { className: "token punctuation", children: "," }), " nrow ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "2" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " heights ", _jsx(_components.span, { className: "token operator", children: "=" }), " c", _jsx(_components.span, { className: "token punctuation", children: "(" }), _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# \"gridExtra\"包可以实现上下分图" }), "\n"] })] }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.3.png", alt: "", className: "internal" }), "\n", _jsxs(_components.h1, { id: "下游分析", children: [_jsx(_components.a, { className: "heading-link", href: "#下游分析", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "下游分析"] }), "\n", _jsxs(_components.h2, { id: "数据标准化scale", children: [_jsx(_components.a, { className: "heading-link", href: "#数据标准化scale", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "数据标准化（scale）"] }), "\n", _jsxs(_components.p, { children: ["实现数据的线性转换(scaling)，也是降维处理之前的标准预处理。主要利用了", _jsx(_components.code, { className: "language-auto", children: "ScaleData" }), "函数，\n之前归一化（normalize）作log处理，它是对所有基因的表达量统一对待的，最后放在了一个起跑线上。但是为了真正找到差异，我们还要基于这个起跑线，考虑不同样本对表达量的影响"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " ScaleData", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " vars.to.regress ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"percent.mt\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# vars.to.regress用于回归的参数，它指定了要从数据中回归掉的变量，这里就是为了移除线粒体细胞等的影响" }), "\n"] }) }) }), "\n", _jsxs(_components.blockquote, { children: ["\n", _jsx(_components.p, { children: "Normalization “normalizes” within the cell for the difference in sequenicng depth / mRNA throughput. 主要着眼于样本的文库大小差异" }), "\n", _jsx(_components.p, { children: "Scaling “normalizes” across the sample for differences in range of variation of expression of genes . 主要着眼于基因的表达分布差异" }), "\n"] }), "\n", _jsxs(_components.h2, { id: "降维", children: [_jsx(_components.a, { className: "heading-link", href: "#降维", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "降维"] }), "\n", _jsx(_components.p, { children: "降维是指通过保留一些比较重要的特征，去除一些冗余的特征，减少数据特征的维度。而特征的重要性取决于该特征能够表达多少数据集的信息，也取决于使用什么方法进行降维。一般情况会先使用线性的降维方法再使用非线性的降维方法，通过结果去判断哪种方法比较合适。" }), "\n", _jsxs(_components.h3, { id: "pca-principal-component-analysis线性降维", children: [_jsx(_components.a, { className: "heading-link", href: "#pca-principal-component-analysis线性降维", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "PCA (principal component analysis)线性降维："] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " RunPCA", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " features ", _jsx(_components.span, { className: "token operator", children: "=" }), " VariableFeatures", _jsx(_components.span, { className: "token punctuation", children: "(" }), "object ", _jsx(_components.span, { className: "token operator", children: "=" }), " pbmc", _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["print", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token punctuation", children: "[" }), _jsx(_components.span, { className: "token string", children: "\"pca\"" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), _jsx(_components.span, { className: "token punctuation", children: "]" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "3" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " nfeatures ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "5" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" }), _jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token operator", children: "-" }), _jsx(_components.span, { className: "token operator", children: "-" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["PC_ ", _jsx(_components.span, { className: "token number", children: "1" }), " \n"] }), _jsxs(_components.span, { className: "code-line", children: ["Positive", _jsx(_components.span, { className: "token operator", children: ":" }), " CST3", _jsx(_components.span, { className: "token punctuation", children: "," }), " TYROBP", _jsx(_components.span, { className: "token punctuation", children: "," }), " LST1", _jsx(_components.span, { className: "token punctuation", children: "," }), " AIF1", _jsx(_components.span, { className: "token punctuation", children: "," }), " FTL \n"] }), _jsxs(_components.span, { className: "code-line", children: ["Negative", _jsx(_components.span, { className: "token operator", children: ":" }), " MALAT1", _jsx(_components.span, { className: "token punctuation", children: "," }), " LTB", _jsx(_components.span, { className: "token punctuation", children: "," }), " IL32", _jsx(_components.span, { className: "token punctuation", children: "," }), " IL7R", _jsx(_components.span, { className: "token punctuation", children: "," }), " CD2 \n"] }), _jsxs(_components.span, { className: "code-line", children: ["PC_ ", _jsx(_components.span, { className: "token number", children: "2" }), " \n"] }), _jsxs(_components.span, { className: "code-line", children: ["Positive", _jsx(_components.span, { className: "token operator", children: ":" }), " CD79A", _jsx(_components.span, { className: "token punctuation", children: "," }), " MS4A1", _jsx(_components.span, { className: "token punctuation", children: "," }), " TCL1A", _jsx(_components.span, { className: "token punctuation", children: "," }), " HLA", _jsx(_components.span, { className: "token operator", children: "-" }), "DQA1", _jsx(_components.span, { className: "token punctuation", children: "," }), " HLA", _jsx(_components.span, { className: "token operator", children: "-" }), "DQB1 \n"] }), _jsxs(_components.span, { className: "code-line", children: ["Negative", _jsx(_components.span, { className: "token operator", children: ":" }), " NKG7", _jsx(_components.span, { className: "token punctuation", children: "," }), " PRF1", _jsx(_components.span, { className: "token punctuation", children: "," }), " CST7", _jsx(_components.span, { className: "token punctuation", children: "," }), " GZMA", _jsx(_components.span, { className: "token punctuation", children: "," }), " GZMB \n"] }), _jsxs(_components.span, { className: "code-line", children: ["PC_ ", _jsx(_components.span, { className: "token number", children: "3" }), " \n"] }), _jsxs(_components.span, { className: "code-line", children: ["Positive", _jsx(_components.span, { className: "token operator", children: ":" }), " HLA", _jsx(_components.span, { className: "token operator", children: "-" }), "DQA1", _jsx(_components.span, { className: "token punctuation", children: "," }), " CD79A", _jsx(_components.span, { className: "token punctuation", children: "," }), " CD79B", _jsx(_components.span, { className: "token punctuation", children: "," }), " HLA", _jsx(_components.span, { className: "token operator", children: "-" }), "DQB1", _jsx(_components.span, { className: "token punctuation", children: "," }), " HLA", _jsx(_components.span, { className: "token operator", children: "-" }), "DPA1 \n"] }), _jsxs(_components.span, { className: "code-line", children: ["Negative", _jsx(_components.span, { className: "token operator", children: ":" }), " PPBP", _jsx(_components.span, { className: "token punctuation", children: "," }), " PF4", _jsx(_components.span, { className: "token punctuation", children: "," }), " SDPR", _jsx(_components.span, { className: "token punctuation", children: "," }), " SPARC", _jsx(_components.span, { className: "token punctuation", children: "," }), " GNG11 \n"] })] }) }), "\n", _jsx(_components.p, { children: "默认选择之前鉴定的差异基因（2000个）作为input，但可以使用features进行设置；默认分析的主成分数量是5个" }), "\n", _jsx(_components.p, { children: "上述print结果可以可视化：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["VizDimLoadings", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "2" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "3" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " reduction ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"pca\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# dims指定任意两个相邻PC" }), "\n"] }) }) }), "\n", _jsx(_components.p, { children: "提取两个主成分（默认前两个，当然可以修改dims选项）绘制散点图：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["DimPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " reduction ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"pca\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.4.png", alt: "", className: "internal" }), "\n", _jsx(_components.p, { children: "接下来是热力图，每个细胞和基因都根据PCA结果得分进行了排序，默认画前30个基因（nfeatures设置），1个主成分(dims设置)，细胞数没有默认值（cells设置），balanced表示正负得分的基因各占一半。" }), "\n", _jsx(_components.p, { children: "它可以查看数据中异质性的主要来源，并可以确定哪些PCA维度可以用于下一步的下游分析。细胞和特征根据PCA的分数进行排序：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["DimHeatmap", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " cells ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "500" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " balanced ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token boolean", children: "TRUE" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.5.png", alt: "", className: "internal" }), "\n", _jsx(_components.p, { children: "降维后怎么选择合适数量的主成分来代表整个数据集？最简单的方法就是以下代码作图，并查看拐点：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["ElbowPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) }), "\n", _jsx(_components.p, { children: "或者：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " JackStraw", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " num.replicate ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "100" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " ScoreJackStraw", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "20" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["JackStrawPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "20" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] })] }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.6.png", alt: "", className: "internal" }), "\n", _jsxs(_components.p, { children: ["第一种方式在PC9-10周围观察到“elbow ”，第二种方式在10-12台PC之后，重要性似乎急剧下降。由此可以用10PC。在选择该参数时候，", _jsx(_components.strong, { children: "尽量偏高一点" }), "。"] }), "\n", _jsxs(_components.h2, { id: "聚类", children: [_jsx(_components.a, { className: "heading-link", href: "#聚类", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "聚类"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FindNeighbors", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "10" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# 上面选了10PC" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FindClusters", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " resolution ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "0.5" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# resolution参数表达聚类的分辨率，值越大得到的cluster越多（看的越清楚），对于3K细胞的单细胞数据0.4-1.2 通常会得到较好的结果。" }), "\n"] })] }) }), "\n", _jsx(_components.p, { children: "此时作图，发现难以通过一个点代表一个细胞的PCA降维可视化图进行呈现，因为在二维或三维的效果下，可视化图会有大量的点（细胞）重叠，仅通过2~3个主成分特征无法很好地将不同亚群之间的数据区分开来：" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["DimPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "2" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " reduction ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"pca\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.7.png", alt: "", className: "internal" }), "\n", _jsxs(_components.h2, { id: "非线性降维umaptsne", children: [_jsx(_components.a, { className: "heading-link", href: "#非线性降维umaptsne", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "非线性降维(UMAP/tSNE)"] }), "\n", _jsx(_components.p, { children: "UMAP与tSNE可以将高维空间的结构特征投影至低维空间（二维、三维）之中：在得到低维分布的具体特征后，计算降维过程中空间位置的信息缺失，优化低维分布的位置信息，保留高维特征信息。这两个降维方式也因此成为了单细胞转录组最主要的可视化方式。" }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " RunUMAP", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "10" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token comment", children: "#可以设置`LABEL=TRUE`或使用LabelClusters函数来帮助标记单个群集" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["umap ", _jsx(_components.span, { className: "token operator", children: "<-" }), " DimPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " reduction ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"umap\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" }), _jsxs(_components.span, { className: "code-line", children: ["pbmc ", _jsx(_components.span, { className: "token operator", children: "<-" }), " RunTSNE", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " dims ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token operator", children: ":" }), _jsx(_components.span, { className: "token number", children: "10" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["tsne ", _jsx(_components.span, { className: "token operator", children: "<-" }), " DimPlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " reduction ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token string", children: "\"tsne\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " label ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token boolean", children: "TRUE" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsx(_components.span, { className: "code-line", children: "\n" }), _jsxs(_components.span, { className: "code-line", children: ["umap ", _jsx(_components.span, { className: "token operator", children: "+" }), " tsne\n"] })] }) }), "\n", _jsx(_components.img, { src: "https://Cafwell.github.io/picx-images-hosting/scRNA/3.8.png", alt: "", className: "internal" }), "\n", _jsxs(_components.h2, { id: "找差异表达基因cluster-biomarker", children: [_jsx(_components.a, { className: "heading-link", href: "#找差异表达基因cluster-biomarker", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "找差异表达基因（cluster biomarker）"] }), "\n", _jsx(_components.p, { children: "目的是根据表达量不同这个特征而分出不同细胞群的基因们（就是找具有生物学意义的HVGs -> 即biomarkers） marker可以理解成标记，就是一看到有这些基因，就说明是这个细胞群体" }), "\n", _jsxs(_components.p, { children: [_jsx(_components.code, { className: "language-auto", children: "FindMarkers()" }), "：比较两个特定cluster之间的基因表达"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token comment", children: "# 找到cluster1中的marker基因" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["cluster1.markers ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FindMarkers", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " ident.", _jsx(_components.span, { className: "token number", children: "1" }), " ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "1" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " min.pct ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "0.25" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["head", _jsx(_components.span, { className: "token punctuation", children: "(" }), "cluster1.markers", _jsx(_components.span, { className: "token punctuation", children: "," }), " n ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "5" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] })] }) }), "\n", _jsx(_components.p, { children: "其中min.pct的意思是：一个基因在任何两群细胞中的占比最低不能低于多少。" }), "\n", _jsxs(_components.p, { children: ["一步到位的办法", _jsx(_components.code, { className: "language-auto", children: "FindAllMarkers()" }), "（对所有cluster都比较一下，并只挑出来正表达marker），它默认会对单独的一个细胞群与其他几群细胞进行比较找到正、负表达biomarker（这里的正负有点上调、下调基因的意思；正marker表示在我这个cluster中表达量高，而在其他的cluster中低）："] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsxs(_components.code, { className: "language-R code-highlight", children: [_jsxs(_components.span, { className: "code-line", children: ["pbmc.markers ", _jsx(_components.span, { className: "token operator", children: "<-" }), " FindAllMarkers", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " only.pos ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token boolean", children: "TRUE" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " min.pct ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "0.25" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " logfc.threshold ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "0.25" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: [_jsx(_components.span, { className: "token comment", children: "# 这一步过滤好好理解(进行了分类、排序、挑前2个)" }), "\n"] }), _jsxs(_components.span, { className: "code-line", children: ["pbmc.markers ", _jsx(_components.span, { className: "token percent-operator operator", children: "%>%" }), " group_by", _jsx(_components.span, { className: "token punctuation", children: "(" }), "cluster", _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token percent-operator operator", children: "%>%" }), " top_n", _jsx(_components.span, { className: "token punctuation", children: "(" }), "n ", _jsx(_components.span, { className: "token operator", children: "=" }), " ", _jsx(_components.span, { className: "token number", children: "2" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " wt ", _jsx(_components.span, { className: "token operator", children: "=" }), " avg_log2FC", _jsx(_components.span, { className: "token punctuation", children: ")" }), " ", _jsx(_components.span, { className: "token comment", children: "# 新版函数用avg_log2FC而不是avg_logFC" }), "\n"] })] }) }), "\n", _jsxs(_components.p, { children: ["速度快的函数：", _jsx(_components.code, { className: "language-auto", children: "marker_cosg <- cosg(sce, groups='all', assay='RNA', slot='data', mu=1, n_genes_user=100)" })] }), "\n", _jsxs(_components.h3, { id: "多种可视化方式", children: [_jsx(_components.a, { className: "heading-link", href: "#多种可视化方式", children: _jsx(_components.svg, { xmlns: "http:www.w3.org/2000/svg", fill: "#ab2b65", viewBox: "0 0 20 20", className: "w-5 h-5", children: _jsx(_components.path, { fillRule: "evenodd", clipRule: "evenodd", d: "M9.493 2.853a.75.75 0 00-1.486-.205L7.545 6H4.198a.75.75 0 000 1.5h3.14l-.69 5H3.302a.75.75 0 000 1.5h3.14l-.435 3.148a.75.75 0 001.486.205L7.955 14h2.986l-.434 3.148a.75.75 0 001.486.205L12.456 14h3.346a.75.75 0 000-1.5h-3.14l.69-5h3.346a.75.75 0 000-1.5h-3.14l.435-3.147a.75.75 0 00-1.486-.205L12.045 6H9.059l.434-3.147zM8.852 7.5l-.69 5h2.986l.69-5H8.852z" }) }) }), "多种可视化方式"] }), "\n", _jsx(_components.pre, { className: "language-r", children: _jsx(_components.code, { className: "language-R code-highlight", children: _jsxs(_components.span, { className: "code-line", children: ["FeaturePlot", _jsx(_components.span, { className: "token punctuation", children: "(" }), "pbmc", _jsx(_components.span, { className: "token punctuation", children: "," }), " features ", _jsx(_components.span, { className: "token operator", children: "=" }), " c", _jsx(_components.span, { className: "token punctuation", children: "(" }), _jsx(_components.span, { className: "token string", children: "\"MS4A1\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"GNLY\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"CD3E\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"CD14\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"FCER1A\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"FCGR3A\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"LYZ\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"PPBP\"" }), _jsx(_components.span, { className: "token punctuation", children: "," }), " ", _jsx(_components.span, { className: "token string", children: "\"CD8A\"" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), _jsx(_components.span, { className: "token punctuation", children: ")" }), "\n"] }) }) })] }); } function MDXContent(props = {}) { const {wrapper: MDXLayout} = { ..._provideComponents(), ...props.components }; return MDXLayout ? _jsx(MDXLayout, { ...props, children: _jsx(_createMdxContent, { ...props }) }) : _createMdxContent(props); } return { default: MDXContent };