Microbiome transplant analysis
load("data/data.Rdata")
transplant <- read_tsv("data/transplant.tsv") %>%
mutate(pair=paste0("P",c(1:39)))
sample_subset <- c(paste0(transplant$donor,"DR",sep=""),paste0(transplant$receptor,"T3",sep=""))
# Original sample comparison
transplanto<- transplant %>%
filter(group=="transplant") %>%
mutate(donor=paste0(receptor,"DR",sep="")) %>%
mutate(receptor=paste0(receptor,"T3",sep="")) %>%
mutate(group="original")
transplant1 <- transplant %>%
filter(group=="transplant") %>%
mutate(donor=paste0(donor,"DR",sep="")) %>%
mutate(receptor=paste0(receptor,"T1",sep="")) %>%
mutate(group=ifelse(group=="transplant","FMT1",group))
transplant2 <- transplant %>%
filter(group=="transplant") %>%
mutate(donor=paste0(donor,"DR",sep="")) %>%
mutate(receptor=paste0(receptor,"T2",sep="")) %>%
mutate(group="FMT2")
transplant3 <- transplant %>%
filter(group=="transplant") %>%
mutate(donor=paste0(donor,"DR",sep="")) %>%
mutate(receptor=paste0(receptor,"T3",sep="")) %>%
mutate(group="FMT3")
anti <- transplant %>%
filter(group=="transplant") %>%
mutate(donor=paste0(donor,"DR",sep="")) %>%
mutate(receptor=paste0(receptor,"AN",sep="")) %>%
mutate(group="anti")
transplant_all <- transplant %>%
filter(group=="control") %>%
mutate(donor=paste0(receptor,"DR",sep="")) %>%
mutate(receptor=paste0(receptor,"T3",sep="")) %>%
bind_rows(transplanto) %>%
bind_rows(transplant1) %>%
bind_rows(transplant2) %>%
bind_rows(transplant3) %>%
bind_rows(anti)
sample_subset2 <- c(paste0(transplant$donor,"DR",sep=""),
paste0(transplant$receptor,"DR",sep=""),
paste0(transplant$receptor,"T1",sep=""),
paste0(transplant$receptor,"T2",sep=""),
paste0(transplant$receptor,"T3",sep=""),
paste0(transplant$receptor,"AN",sep="")) %>%
unique()
#Calculate Hill numbers
beta_neutral_table <- genome_counts %>%
select(all_of(c("genome",sample_subset2))) %>%
column_to_rownames(var="genome") %>%
hillpair(.,q=1, metric="C", out = "pair")
beta_neutral_table %>%
inner_join(transplant_all,by=join_by(first==donor,second==receptor)) %>%
mutate(group=factor(group,levels=c("control","original","anti","FMT1","FMT2","FMT3"))) %>%
group_by(group) %>%
summarise(mean=mean(C), sd=sd(C)) %>%
tt()
| group |
mean |
sd |
| control |
0.1667197 |
0.073536006 |
| original |
0.4944570 |
0.216215379 |
| anti |
0.9984575 |
0.008306813 |
| FMT1 |
0.6225804 |
0.214404391 |
| FMT2 |
0.5475272 |
0.224134831 |
| FMT3 |
0.5174177 |
0.230051302 |
beta_neutral_table %>%
inner_join(transplant_all,by=join_by(first==donor,second==receptor)) %>%
mutate(group=factor(group,levels=c("control","original","anti","FMT1","FMT2","FMT3"))) %>%
ggplot(aes(x=group,y=C,group=group)) +
geom_boxplot() +
theme_minimal()
#Calculate Hill numbers
beta_neutral <- genome_counts %>%
select(all_of(c("genome",sample_subset))) %>%
column_to_rownames(var="genome") %>%
hillpair(.,q=1, metric="C")
beta_neutral %>%
vegan::metaMDS(., trymax = 500, k = 2, verbosity = FALSE, trace=FALSE) %>%
vegan::scores() %>%
as_tibble(., rownames = "sample") %>%
dplyr::left_join(sample_metadata, by = join_by(sample == sample)) %>%
dplyr::full_join(transplant %>% select(donor,pair), by = join_by(animal == donor)) %>%
dplyr::full_join(transplant %>% select(receptor,pair), by = join_by(animal == receptor)) %>%
mutate(pair.x=ifelse(treatment=="DR",pair.x,NA)) %>%
mutate(pair.y=ifelse(treatment=="T3",pair.y,NA)) %>%
rowwise() %>%
mutate(pair = list(c(pair.x, pair.y))) %>%
unnest(pair) %>%
filter(!is.na(pair)) %>%
ggplot(aes(x = NMDS1, y = NMDS2, color=treatment, shape=group, group = pair)) +
scale_shape_manual(name="Group",
breaks=c("invariable","variable"),
labels=c("Control","Transplant"),
values=c(17,19)) +
scale_color_manual(name="Sample",
breaks=c("DR","T3"),
labels=c("Donor","Receptor"),
values=c("#d8b8a3","#bcdee1")) +
scale_linetype_manual(name="Group",
breaks=c("invariable","variable"),
labels=c("Control","Transplant"),
values=c("dashed","solid")) +
geom_point(size = 4) +
geom_path(aes(linetype=group),color="#999999", alpha=0.5, arrow = arrow(length=unit(0.30,"cm"), ends="last", type = "closed")) +
theme_classic()
