UC_vs_ileal_or_colonic.Rmd

---
title: "R Notebook"
output: html_notebook
---

This is an [R Markdown](http://rmarkdown.rstudio.com) Notebook. When you execute code within the notebook, the results appear beneath the code. 

Try executing this chunk by clicking the *Run* button within the chunk or by placing your cursor inside it and pressing *Cmd+Shift+Enter*. 

```{r}
library(Hmisc)
library(psych)
library(gplots)
library(RColorBrewer)
library(corrplot)
library(matrixTests)
library(ggplot2)
library(dplyr)
```
import data
```{r}
all_data = read.delim(pipe("pbpaste"), row.names = 1)
```
subset diff sets
```{r}
ileal = subset(all_data, Active_Location == "ileal")
colonic = subset(all_data, Active_Location == "colonic")
UC = subset(all_data, Diagnosis == "UC")

ileal = ileal[,3:ncol(ileal)]
colonic = colonic[,3:ncol(colonic)]
UC = UC[,3:ncol(UC)]
```

do stats and add fold change
```{r}
IVUC = col_t_welch(ileal, UC)
CVUC = col_t_welch(colonic, UC)

IVUC$padjust = p.adjust(IVUC$pvalue, method = "fdr")
CVUC$padjust = p.adjust(CVUC$pvalue, method = "fdr")

IVUC$FC = log2(IVUC$mean.x/IVUC$mean.y)
CVUC$FC = log2(CVUC$mean.x/CVUC$mean.y)

```


```{r}
write.csv(IVUC, "ileal_v_UC_sig.csv")
write.csv(CVUC, "colonic_v_UC_sig.csv")
```

```{r}
col_il_UC = subset(all_data, Active_Location == "colonic" | Active_Location == "ileal" | Diagnosis == "UC")
col_il_UC_new = col_il_UC %>%
  mutate(Active_Location = replace(Active_Location, Active_Location == "UC", 3)) %>%
  mutate(Active_Location = replace(Active_Location, Active_Location == "colonic", 2)) %>%
  mutate(Active_Location = replace(Active_Location, Active_Location == "ileal", 1))

samp_key = col_il_UC_new$Active_Location

col_il_UC_mat = as.matrix(col_il_UC[,3:ncol(col_il_UC)])
```

```{r}
col_il_UC_stats = col_oneway_welch(col_il_UC_mat, samp_key)
```
```{r}
col_il_UC_stats$padjust = p.adjust(col_il_UC_stats$pvalue, method = "fdr")
```

```{r}
col_il_UC_sig_only = subset(col_il_UC_stats, pvalue < 0.002)
sig_feats_only = rownames(col_il_UC_sig_only)
```

```{r}
col_il_UC_sig_data = subset(col_il_UC_mat, select = sig_feats_only)
```

```{r}
hcolors = colorRampPalette(c("cyan", "black", "yellow")) (n=20)
heatmap.2(col_il_UC_sig_data, trace = "none", scale = "column", col = hcolors, cexRow = 0.3, cexCol = 0.3, RowSideColors = samp_key, Rowv = F)
```