CRC_analysis.Rmd

---
title: "CRC_analysis"
output: html_document
date: "2023-07-12"
---
---
title: "SpaceANOVA"
author: Souvik Seal
output: rmarkdown::github_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#",
  out.width = "100%",
  messages = FALSE,
  warnings = FALSE
)
```

This is an R package implementing the proposed method from the paper, "SpaceANOVA: Spatial co-occurrence analysis of cell types in multiplex imaging data using point
process and functional ANOVA". It can be used to study differential spatial co-occurrence of pairs of cell types across multiple groups, such as case/control, in multiplex imaging datasets. Here, we show how this package was used on the CRC dataset from yhe main manuscript.

## Loading the package

We install and load the developmental version of SpaceANOVA from GitHub.

```{r loading packages, message=FALSE}

suppressWarnings(devtools::install_github('sealx017/SpaceANOVA', quiet = TRUE))
require(SpaceANOVA)

require(spatstat)
require(tidyr)
require(dplyr)
require(fda.usc)
require(ggplot2)
require(cowplot)
require(gridExtra)

```

## Loading the example dataset

Next, we import the colorectal cancer (CRC) MIF dataset [1]. There are two groups of subjects: one group (18 subjects) with the “Crohn’slike reaction” (CLR) tumor invasive front, while the other group (17 subjects) is defined by the presence of diffuse inflammatory infiltration (DII).

```{r loading the marker expression, out.width = "50%"}
data("MIF_CRC")
knitr::kable(head(MIF_CRC), format="markdown")
MIF_CRC %>% group_by(Group, ID) %>% reframe(n = n()) # check group and subject IDs, and respective cell counts (spread across multiple images)

```

## Computing the summary functions and performing association analysis

We keep the range of radius r was kept between [0, 100] with a step size of 10 which is higher than what we considered in the other two real data analyses because the number of cells is much lower in this dataset, and a finer grid of r would result in noisier and less smooth g-functions. For a similar reason and the tissue missing-ness not being a highlighted issue by Schutch et al. (2020) [1], we consider unadjusted
versions of SpaceANOVA Univ. and SpaceANOVA Mult. (perm = FALSE) in this case.

```{r Computing spatial summary functions, and performing both univariate and multivariate differential co-occurrence analysis, echo = T, results = T}

Final_result = All_in_one(data = MIF_CRC, fixed_r = seq(0, 100, by = 10), Summary_function = "g", Hard_ths = 20, homogeneous = TRUE, interaction_adjustment = TRUE, perm = TRUE, cores = 8)

```

## Extracting p-values 

Extract the p-values of the univariate and multivariate tests: SpaceANOVA Univ. and SpaceANOVA Mult. and print.

```{r P-value extraction}
p_res = p_extract(Final_result)
Univ_p = p_res[[1]]
Mult_p = p_res[[2]]
print(Univ_p)

```

## Heatmap of -log10 of p-values 


Display the -log10 of the p-values using heatmap.

```{r P-value visualization,  out.width="50%", out.height = "100%"}
Plot.heatmap(Univ_p, main = "SpaceANOVA Univ.")
```

## Visualizing summary functions

Next, we can check the summary functions corresponding to those pairs of cell types whose p-values turned out to be significant. Here, we visualize the g-functions of pairs: (b_cells, tumor_cells) and (vasculature, tumor_cells). For every pair, the first panel shows subject-level mean functions (averaged across the images of a subject), while the second panel shows image-level summary functions. We do not display the point-wise F-values here, i.e., Fvals = NULL but one can specify Fvals = Pointwise_Fvals to check the associated pointwise F values. 

### Pair: (b_cells, tumor_cells)

We notice that the spatial co-occurrence of b_cells and tumor cells is negative in the CLR group for smaller values of r, while in DII the co-occurrence is always negative. 

```{r plotting summary functions 1,  out.width="100%", out.height = "80%"}
Pointwise_Fvals = Final_result[[1]][[2]]
Functional_results = Final_result[[2]]

# Pair: (b_cells, tumor_cells)
Plot.functions(Functional_results, pair = c("b_cells", "tumor_cells"), Fvals = NULL)

```

### Pair: (vasculature, tumor_cells)


```{r plotting summary functions 2,  out.width="100%", out.height = "80%"}
# Pair: (beta, Tc)
Plot.functions(Functional_results, pair = c("vasculature", "tumor_cells"), Fvals = NULL)

```

## Visualizing cellular organization

We can visualize cellular organization in different images of every subject. We create a reduced dataset first to focus only on b_cells, tumor_cells and vasculature.

```{r plotting celluar organization,  out.width="100%", out.height = "500px"}

MIF_CRC_reduced = MIF_CRC
MIF_CRC_reduced$cellType = as.character(MIF_CRC_reduced$cellType)
MIF_CRC_reduced$cellType = ifelse(MIF_CRC_reduced$cellType %in% c("b_cells", "tumor_cells", "vasculature"), MIF_CRC_reduced$cellType, "Others")

table(MIF_CRC_reduced$cellType)
palette = c("red", "grey", "purple", "green4") #assign colors to cell types 

Plot.cellTypes(data = MIF_CRC_reduced, ID = "28", imageID = c("reg055_A", "reg056_A"), palette = palette)
Plot.cellTypes(data = MIF_CRC_reduced, ID = "3",  imageID = c("reg005_A", "reg006_A"), palette = palette)

```


## References

1\) Schürch, C. M., Bhate, S. S., Barlow, G. L., Phillips, D. J., Noti, L., Zlobec, I., ... & Nolan, G. P. (2020). Coordinated cellular neighborhoods orchestrate antitumoral immunity at the colorectal cancer invasive front. Cell, 182(5), 1341-1359.

2\) Baddeley, A., Rubak, E., and Turner, R. (2015). Spatial point patterns: methodology and applications with R. CRC press.

3\) Zhang, J.-T. (2013). Analysis of variance for functional data. CRC press.