Skip to content

Commit

Permalink
formatting changes
Browse files Browse the repository at this point in the history
  • Loading branch information
@jiajic
jiajic committed Oct 28, 2024
1 parent a2e063a commit ae8fad2
Show file tree
Hide file tree
Showing 2 changed files with 85 additions and 29 deletions.
102 changes: 77 additions & 25 deletions vignettes/object_creation.Rmd
View file
Original file line number Diff line number Diff line change
Expand Up @@ -44,12 +44,20 @@ With Giotto, the minimal requirements for a spatial dataset are either:

Single cell datasets (shown at the bottom) can also be worked using Giotto, and only need a matrix without the need for paired spatial locations.

# Create from matrix and spatial locations


# Create from:

## Matrix and spatial locations

<details>
<summary>expand</summary>


This example will be shown using an osmFISH dataset.
To download this data, please ensure that [wget](https://www.gnu.org/software/wget/?) is installed locally.

## Download data to use
### Download data to use
```{r, eval=FALSE}
# Specify path from which data may be retrieved/stored
data_path <- "/path/to/data/"
Expand All @@ -65,8 +73,7 @@ GiottoData::getSpatialDataset(
)
```


## Read data into Giotto Suite
### Read data into Giotto Suite

Aggregate data (matrix and spatial locations -based data) can be read in using `createGiottoObject()`.
There are two examples. Data formatting guidelines are shown below this code block.
Expand Down Expand Up @@ -127,7 +134,7 @@ spatial locations ----------------
Use objHistory() to see steps and params used
```

## Example plot
### Example plot
```{r, eval=FALSE}
spatFeatPlot2D(minimum_gobject1,
feats = c("Sox10", "Gfap"),
Expand All @@ -143,7 +150,7 @@ knitr::include_graphics("images/object_creation/1_spatplot_aggregate.png")
```


## Data formatting notes
### Data formatting notes

**Expression file formatting**

Expand Down Expand Up @@ -205,13 +212,22 @@ spatial locations ----------------
Use objHistory() to see steps and params used
```

</details>






# Create from feature detection (points) and cell annotations (polygons)
## Feature detection (points) and cell annotations (polygons)

<details>
<summary>expand</summary>


You can also make giotto objects starting from raw spatial feature information and annotations that give them spatial context. This will be shown using the vizgen MERSCOPE mini object's raw data available from GiottoData

## Get filepaths from GiottoData
### Get filepaths from GiottoData

```{r, eval=FALSE}
# function to get a filepath from GiottoData
Expand All @@ -227,7 +243,7 @@ mini_viz_tx_path <- mini_viz_raw("vizgen_transcripts.gz")
```


## Read into Giotto subobjects and assemble into giotto object
### Read into Giotto subobjects and assemble into giotto object

Polygons can be read into Giotto Suite in multiple ways.

Expand Down Expand Up @@ -368,7 +384,7 @@ knitr::include_graphics("images/object_creation/3_viz_gpoints.png")
```


## Checking spatial alignment
### Checking spatial alignment

Care should always be taken when assembling a spatial dataset to make sure that
the spatial information is spatially aligned. We can check this by plotting the subobjects.
Expand Down Expand Up @@ -401,7 +417,7 @@ knitr::include_graphics("images/object_creation/5_align_ROI.png")
```


## Creating the subcellular Giotto object
### Creating the subcellular Giotto object

Datasets with raw spatial features and polygon information are created using `createGiottoObjectSubcellular()`.

Expand All @@ -425,7 +441,7 @@ features : rna
Use objHistory() to see steps and params used
```

## Spatially aggregate values
### Spatially aggregate values

```{r, eval=FALSE}
# calculate centroids
Expand All @@ -435,7 +451,7 @@ mini_viz <- calculateOverlap(mini_viz)
mini_viz <- overlapToMatrix(mini_viz)
```

## Example plot
### Example plot

```{r, eval=FALSE}
spatFeatPlot2D(
Expand All @@ -452,11 +468,25 @@ spatFeatPlot2D(
knitr::include_graphics("images/object_creation/6_example_featplot_sub.png")
```

# Create from staining images and polygons
</details>









## Staining images and polygons

<details>
<summary>expand</summary>


The IMC data to run this tutorial can be found in the *imcdatasets* package. We will be using one of the smaller example datasets. These IMC outputs provide multilayer images in the format of *cytomapper* `CytoImageList` collections of *EBImage* `Image` objects, which can be easily coerced to simple matrices.

## Download data to use
### Download data to use

The package *exactextractr* is also used due to its speed with image data extractions.

Expand All @@ -483,7 +513,7 @@ img_mat <- lapply(
channels <- cytomapper::channelNames(imc_imgs)
```

## Read data into Giotto Suite
### Read data into Giotto Suite

```{r, eval=FALSE}
# create the mask polys
Expand Down Expand Up @@ -556,7 +586,7 @@ plot(imglist[["H3"]], max_intensity = 30)
knitr::include_graphics("images/object_creation/8_img_H3.png")
```

## Checking spatial alignment
### Checking spatial alignment

```{r, eval=FALSE}
plot(imglist[["H3"]], max_intensity = 30)
Expand All @@ -567,7 +597,7 @@ plot(mask_poly, add = TRUE, border = "cyan", lwd = 0.15)
knitr::include_graphics("images/object_creation/9_align2.png")
```

## Creating the giotto object
### Creating the giotto object

```{r, eval=FALSE}
imc <- createGiottoObjectSubcellular(
Expand All @@ -591,7 +621,7 @@ images : 40 items...
Use objHistory() to see steps and params used
```

## Spatially aggregate values
### Spatially aggregate values

```{r, eval=FALSE}
# calculate centroids
Expand All @@ -601,7 +631,7 @@ imc <- calculateOverlap(imc, image_names = channels, name_overlap = "protein")
imc <- overlapToMatrix(imc, type = "intensity", feat_info = "protein")
```

## Example plot
### Example plot

```{r, eval=FALSE}
spatFeatPlot2D(
Expand All @@ -621,8 +651,16 @@ knitr::include_graphics("images/object_creation/10_example_plot_imc.png")
**Note:** If you notice a single extremely bright center spot, then there was likely a background polygon that was not removed.


</details>


# Create piecewise



## Create piecewise

<details>
<summary>expand</summary>

Giotto objects can also be created in a piecewise manner as long as all inputs
are converted to Giotto subobject formats first. `giottoPoints` and `giottoPolygon`
Expand Down Expand Up @@ -651,11 +689,18 @@ Use objHistory() to see steps and params used

This is essentially the same object as the one created through `createGiottoObjectSubcellular()` with the vizgen MERSCOPE data earlier.

</details>



# Create a single cell analysis Giotto object

## Download data to use
## Single cell data

<details>
<summary>expand</summary>


### Download data to use

The example dataset used here is from 10X [link](https://www.10xgenomics.com/datasets/Human_PBMCs_Next_GEM_Flex_GEM-X_Flex_Comparison)

Expand All @@ -669,13 +714,13 @@ save_path <- file.path(data_path, "pbmc_matrix.h5")
utils::download.file(link, destfile = save_path, method = "wget")
```

## Read the data
### Read the data

```{r, eval=FALSE}
pbmc_mat <- get10Xmatrix_h5(save_path)
```

## Load into Giotto as a single cell analysis
### Load into Giotto as a single cell analysis

```{r, eval=FALSE}
pbmc <- createGiottoObject(expression = pbmc_mat)
Expand All @@ -698,6 +743,13 @@ spatial locations ----------------
Use objHistory() to see steps and params used
```

</details>







# Convenience functions for specific platforms

Expand Down
12 changes: 8 additions & 4 deletions vignettes/split_join.Rmd
View file
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,10 @@ This example shows working with a 10x Visium
[mouse TMA](https://www.10xgenomics.com/datasets/mouse-tissue-microarray-in-3x3-layout-with-2-mm-edge-to-edge-spacing-ffpe-2-standard)
dataset.

# Setup


*Check Giotto installation*

```{r, eval=FALSE}
# Ensure Giotto Suite is installed.
if(!"Giotto" %in% installed.packages()) {
Expand All @@ -36,8 +39,9 @@ if(!genv_exists){
}
```

# Setup example TMA dataset

# Download the Data
## Download the Data

```{r, eval=FALSE}
## provide path to a folder to save the example data to
Expand All @@ -58,7 +62,7 @@ spat_dir <- file.path(data_path, "spatial")
```


# Create Giotto Object
## Create Giotto Object

```{r, eval=FALSE}
library(Giotto)
Expand Down Expand Up @@ -91,7 +95,7 @@ spatPlot2D(tma, show_image = TRUE, point_size = 1.5, point_alpha = 0.7)
knitr::include_graphics("images/split_join/1_spatplot.png")
```

# Detect individual cores
## Detect individual TMA cores

By building a spatial network, we can find out which data points are spatially
contiguous by looking checking spatial distance, breaking ones that are too
Expand Down

0 comments on commit ae8fad2

Please sign in to comment.