How can I add colored tile under the tip label?

[ARise-fox]

I learned that use geom_fruit and geom_tile can add bars to the circular tree graph, but when my tip nodes do not have the same x value, some tiles cannot be shown on the plot with the warning message

Warning message: Removed 1 rows containing missing values (geom_tile()).

and here is my output figure:

It's clear that node "Microcoleus_asticus_IPMA8" has no colored shading.

I want to get a figure like these followings, having colored ring under species the tip label, and grouped by phylum.

Could you please offer me some help? How can I make figures like these using ggtree and ggtree extra?
Here is my code and origin file:

tree_file:
id50.txt
csv_file:
id50_items.csv

tree=read.tree(tree_file)
data=fortify(tree)
data_csv <- read.csv(csv_file, check.names = FALSE) #记录物种分类信息
data_csv[data_csv == ''] <- NA
# la:"domain"  "kingdom" "phylum"  "class"   "order"   "family"  "genus"   "species"
la <- names(data_csv)[-1]
mer_col <- c(names(data),names(data_csv))
full.df <- data.frame()
for (i in la) {
  # la的第一个，也就是data_csv的第二个，遍历到names(data_csv) = i 的index
  j1<-left_join(data, data_csv, by=c("label"= i))
  na.omit(j1) ->j2
  j2$input_term <- j2$label 
  # 加入名称为i的列
  j2$temp = j2$label 
  colnames(j2)[colnames(j2) == "temp"] <- i 
  # 按顺序排列
  j2 <- j2[,mer_col]
  # 该level后全设NA
  if(i != "species"){
    j2[,c((which(mer_col == i)+1):ncol(j2))] <- NA
  }
  j2 <- distinct(j2,.keep_all = T) 
  rbind(full.df,j2) %>%  distinct(.keep_all = T) -> full.df 
}
# 未匹配情况，包含根结点和非标准分类元
nomap.df <- subset(data, node %in% data$node & !node %in% full.df$node )
# 合并未匹配情况
left_join(nomap.df,data_csv,by = c("label" = "input_term"))%>%
  mutate(input_term=label)%>%
  rbind(full.df,.) -> full.df

# draw circular graph
cirgraph <- ggtree(tree, layout = "circular") + 
  # geom_hilight(data=pdata, aes(node=node, fill=phylum),type = "auto")+
  geom_tippoint(data = data, size=1)+
  # geom_tiplab(data=data, size = 2,align=T, linetype=3, linesize=0.5, hjust=-0.02, aes(color = phylum))+
  geom_nodepoint(size=0.5, color="steelblue")+
  xlim(NA, max(data$x)*1.3)

# dataframe to add annotation
dat2 <- full.df[,c("node","x","label", "phylum")]

p3 <- cirgraph +
  geom_fruit(
    data = dat2,
    geom = geom_tile,
    mapping=aes(x,y=label , fill=phylum),
    alpha = 0.4,
    offset=0.1,   # The distance between layers, default is 0.03 of x range of tree.
    pwidth=0.4, # width of the layer, default is 0.2 of x range of tree.
    # the `position` parameter was set to `identity` to add the points on the tip nodes.
    # position=position_identityx(hexpand = 8),
  ) +
  geom_tiplab(size = 1.4,align=T, linetype=3, linesize=0.5, hjust=-0.02, offset = 2)

[ARise-fox]

Also， I can't set the param offset and pwidth too high like offset=0.2, pwidth=0.2 , or else there will be a message:
Warning message:
Removed 50 rows containing missing values (geom_tile()).
I guess that indicate the tile elements is drawn out of layer scope.

[xiangpin]

Also， I can't set the param offset and pwidth too high like offset=0.2, pwidth=0.2 , or else there will be a message:
Warning message:
Removed 50 rows containing missing values (geom_tile()).
I guess that indicate the tile elements is drawn out of layer scope.

This is because you use xlim, the layer was removed, set a bigger value can solve the problem, or don't use it when using circular layout. In addition, when x only have one unique value, and you are using geom_tile, please use width argument of geom_tile to replace pwidth, you can refer to the answer.

It's clear that node "Microcoleus_asticus_IPMA8" has no colored shading.

This is because the tip.label of tree is Microcoleus_asticus_IPMA8, but the dat you provided is others.

You can refer to the following codes.

library(ggtree)
library(ggtreeExtra)

tr <- read.tree('./id50.txt')

dat <- read.csv('./id50_items.csv')


# You should provide a column the same with tip.labels of tree
# You can find the input_term does not contain 's__' but
# The tip.label contains except Microcoleus_asticus_IPMA8.
dat$input_term
tr$tip.label <- gsub("s__", "", tr$tip.label)

# Then you can use left_join to integrate the dat.
tr %<>% dplyr::left_join(dat, by=c('label'='input_term'))

p <- ggtree(tr, layout='fan', open.angle=0) +
     geom_tiplab(align = TRUE, size = 3)

# Since the dat had been integrated to the tree, you don't
# need provide again, and the y of mapping also don't need
# specify.
p <- p + geom_fruit(
        geom=geom_tile,
        mapping=aes(fill=phylum),
        alpha=.3,
        width=10,
        offset=.55
     )
p       

[ARise-fox]

Thank you very much! The problem that had been bothering me for a long time was finally solved.