问题描述：

在对自己跑出来的空间转录组数据用Seurat V3.2进行分析时取子集遇到一个问题，我不知道该如何选取像素点坐标，以及用官网给的字段会出现报错的情况，于是我在Github上作者大神提问，不出意外，下面也有一个跟我遇到的情况类似的。

github issue地址： https://github.com/satijalab/seurat/issues/2454

提问原文（biejiaodeyingwen）

Hi~I meet some problem in Subset out anatomical regions

I use the dataset: Mouse Brain Section (Coronal), and run sapceranger successfully,and get clustering_visualization img.

#Subset out anatomical regions
cortex <- subset(brain, idents = c(0, 1, 2, 5, 10, 11, 12 ))
# now remove additional cells, use SpatialDimPlots to visualize what to remove
# SpatialDimPlot(cortex,cells.highlight = WhichCells(cortex, expression = image_imagerow > 400 |
# image_imagecol < 150))
cortex <- subset(cortex, imgine_imagerow > 400 | image_imagecol < 150, invert = TRUE)
cortex <- subset(cortex, image_imagerow > 275 & image_imagecol > 370, invert = TRUE)
cortex <- subset(cortex, image_imagerow > 250 & image_imagecol > 440, invert = TRUE)

then, the error occur:

> #Subset out anatomical regions
> cortex <- subset(brain, idents = c(0, 1, 2, 5, 10, 11, 12 ))
> # now remove additional cells, use SpatialDimPlots to visualize what to remove
> # SpatialDimPlot(cortex,cells.highlight = WhichCells(cortex, expression = image_imagerow > 400 |
> # image_imagecol < 150))
> cortex <- subset(cortex, imgine_imagerow > 400 | image_imagecol < 150, invert = TRUE)
Error in FetchData(object = object, vars = expr.char[vars.use], cells = cells,  : 
  None of the requested variables were found: 
> cortex <- subset(cortex, image_imagerow > 275 & image_imagecol > 370, invert = TRUE)
Error in FetchData(object = object, vars = expr.char[vars.use], cells = cells,  : 
  None of the requested variables were found: 
> cortex <- subset(cortex, image_imagerow > 250 & image_imagecol > 440, invert = TRUE)
Error in FetchData(object = object, vars = expr.char[vars.use], cells = cells,  : 
  None of the requested variables were found:

And I have try to use the solution: #1212 (comment) for subsetting on features that have a dash in their name. But it doesn't work.

So,I wonder the structure of seurat may be the key to solve the problem.
I tape:

cortex@images

and saw：

attr(,"class")
[1] "scalefactors"

Slot "coordinates":
                   tissue row col imagerow imagecol
AAACAATCTACTAGCA-1      1   3  43     1611     4170
AAACACCAATAACTGC-1      1  59  19     8315     2519
AAACAGAGCGACTCCT-1      1  14  94     2927     7679

I try change the imgine_imagerow into imagerow, and change the image_imagecol into the imagecol,but it doesn't work again.

And another question is : How can I further segment based on exact positions?

Because I don't know exactly about the coordinates of cortex.How the example in official website actually do in this step?

thanks!

下面有一个小哥给我的问题点了个赞，居然有些小开心。

作者的回答

Hi,

Please try replacing image_imagerow and image_imagecol with anterior1_imagerow and anterior1_imagecol .

This is still under active development but the do.identify parameter of SpatialDimPlot should allow you to select cells manually or at least give you a sense of where you could start to draw some cutoffs.

我的思考

感觉取子集这部分，有点像旧的空间转录组的流程的ST viewer这步生成坐标映射关系表的感觉
作者也说了，这个版本还在开发中，需要但是可以尝试用下他的输入参数

最近暂时还没有空尝试，不过出于Github的社交礼仪，还是想先对作者表示感谢~

10X空间转录组（七）思考新版Seurat V3.2作者在Github给予的回答

问题描述：

提问原文（biejiaodeyingwen）

作者的回答

我的思考