估计阅读时长: 14 分钟https://github.com/rsharp-lang/ggplot 之前在阅读一篇单细胞组学数据分析的文献,觉得在文献之中有一些三维散点图用于展示降维聚类结果的效果非常的好看。于是自己在R#语言之中的ggplot程序包的2D绘图的功能基础之上,进行了三维图形数据可视化功能的开发。 (A) t-SNE map projecting myeloid cells from BC1-8 patients (all tissues). Cells are colored […]
Recent Posts
Archives
- October 2025 (1)
- August 2025 (3)
- July 2025 (2)
- June 2025 (6)
- May 2025 (3)
- November 2023 (1)
- June 2023 (2)
- May 2023 (2)
- April 2023 (2)
- March 2023 (2)
- February 2023 (1)
- August 2022 (2)
- July 2022 (2)
- June 2022 (5)
- May 2022 (5)
- April 2022 (4)
- March 2022 (3)
- January 2022 (2)
- December 2021 (2)
- November 2021 (2)
- October 2021 (6)
- September 2021 (8)
- August 2021 (8)
- July 2021 (6)
- June 2021 (20)
- May 2021 (10)
Tags
algorithm (33)
bilibili (3)
binary tree (3)
Chromatography (3)
clustering (19)
contour (3)
Darwinism (4)
dataframe (3)
data visualization (23)
dotnet-core (25)
GCModeller (19)
gdi+ (22)
ggplot (14)
graph (14)
heatmap (5)
html (3)
http (4)
image processing (7)
kegg (5)
kmeans (3)
language (7)
linq (3)
linux (8)
machine learning (4)
mass spectrometry (12)
math (19)
MSI (4)
mzkit (19)
network (8)
pathway (4)
pipeline (4)
query (5)
R# (44)
rsharp (23)
scripting (14)
single-cell (6)
sql (3)
symbolic computation (3)
text processing (4)
typescript (3)
ubuntu (4)
uniprot (3)
vb (19)
VisualBasic (50)
webassembly (3)

This clarifies everything perfectly.
其实,你不应该直接跑原始表达矩阵的。因为在原始表达矩阵中,基因的特征数量可能会非常多,做随机森林或者SVM建模就会会非常久。应该先用limma程序包对矩阵筛选一次,例如用log2fc绝对值按照阈值cutoff筛选一次,或者对log2fc绝对值排序后取前1000个特征,得到小一些feature集合的矩阵后再使用这个程序包做机器学习分析。
Thanks for taking the time to create this.
就是随便看看!
c⌒っ゚Д゚)っ救命啊,谢老师,我试了下用这个程序包直接跑转录组矩阵,跑了好久都没有结果