Make flywire_basics visibile

* Function to get some basic information to describe a flywire neuron
* Also flywire_rewrite function now deletes duplicated entries from FAFB tab of matching sheet

NAMESPACE

@@ -86,6 +86,7 @@ export(fafb_set_transmitter)
 export(flow_centrality)
 export(flycircuit_matching_rewrite)
 export(flycircuit_neurons)
+export(flywire_basics)
 export(flywire_contributions)
 export(flywire_failed)
 export(flywire_ids)

R/flywire_googledrive.R

@@ -5,11 +5,14 @@
 #'  without loading the entire, large \code{neuronlist} into memory. You will need access to the hemibrain Google Team Drive and
 #'  have it mounted with Google filestream.The function \code{flywire_neurons_update} can be used to update the available data.
 #'  If you want to flag flywire neurons that should be added to the Google drive, without doing this yourself, you can use
-#'  \code{flywire_request}.
+#'  \code{flywire_request}. The \code{flywire_basics} function will calculate some useful meta-data, namely
+#' a point in the primary neurite tract that can be used as a stable reference for this neuron.
 #'
 #' @param brain the brainspace in which hemibrain neurons have been registered. Defaults to raw voxel space for the FlyWire project.
 #' @param local \code{FALSE} or path. By default (\code{FALSE}) data is read from \code{options()$Drive_hemibrain_data}), but the user can specify an alternative path.
 #' @param mirror logical, whether or not to read neurons that have been mirrored (i.e. flipped to the 'other' brain hemisphere).
+#' @param flywire.neurons a \code{neuronlist} of flywire neurons in FlyWire space. The \code{flywire_basics} function will calculate some useful meta-data, namely
+#' a point in the primary neurite tract that can be used as a stable reference for this neuron.
 #' @param x flywire IDs to update, for the saved Google drive \code{neuronlistfh} objects called with \code{flywire_neurons}.
 #' @param request a neuronlist, matrix of x,y,z position or flywire ID to add to a
 #' \href{https://docs.google.com/spreadsheets/d/1rzG1MuZYacM-vbW7100aK8HeA-BY6dWAVXQ7TB6E2cQ/edit#gid=0}{Google sheet} that records flywire positions
@@ -270,15 +273,15 @@ flywire_nblast_update <- function(x = NULL,
 
 }
 
-
-# hidden
-flywire_basics <- function(x, ...){
-  if(!nat::is.neuronlist(x)){
-    stop("x must be a neuronlist")
+#' @rdname flywire_neurons
+#' @export
+flywire_basics <- function(flywire.neurons, ...){
+  if(!nat::is.neuronlist(flywire.neurons)){
+    stop("flywire.neurons must be a neuronlist")
   }
 
   # Get xyz for primary branch points
-  simp = nat::nlapply(x,nat::simplify_neuron,n=1)
+  simp = nat::nlapply(flywire.neurons,nat::simplify_neuron,n=1)
   branchpoints = sapply(simp, function(y) nat::xyzmatrix(y)[ifelse(length(nat::branchpoints(y)),nat::branchpoints(y),max(nat::endpoints(y))),])
   branchpoints = t(branchpoints)
   flywire.nm.xyz = apply(branchpoints, 1, paste_coords)
@@ -293,12 +296,12 @@ flywire_basics <- function(x, ...){
   #FAFB.xyz = ""
 
   # Add
-  x[,"flywire.id"] =  names(x)
-  x[,"flywire.xyz"] = flywire.xyz
-  x[,"FAFB.xyz"] = FAFB.xyz
-  x[,"dataset"] = "flywire"
-  x[,"id"] = NULL
-  x
+  flywire.neurons[,"flywire.id"] =  names(flywire.neurons)
+  flywire.neurons[,"flywire.xyz"] = flywire.xyz
+  flywire.neurons[,"FAFB.xyz"] = FAFB.xyz
+  flywire.neurons[,"dataset"] = "flywire"
+  flywire.neurons[,"id"] = NULL
+  flywire.neurons
 
 }
 

R/flywire_matching.R

@@ -4,21 +4,30 @@
 #' @export
 flywire_matching_rewrite <- function(flywire.ids = names(flywire_neurons()),
                                      meta = flywire_neurons()[,],
-                                     selected_file  = options()$hemibrainr_matching_gsheet,
+                                     catmaid.update = TRUE,
+                                     selected_file  = options()$hemibrainr_matching_gsheet, # 1_RXfVRw2nVjzk6yXOKiOr_JKwqk2dGSVG_Xe7_v8roU
                                      ...){
   # Get the FAFB matching Google sheet
   gs = hemibrain_match_sheet(sheet = "FAFB", selected_file = selected_file)
   skids = unique(gs$skid)
 
+  # Update side information
+  if(!is.null(meta$side)){
+    side.indices = match(gs$flywire.xyz,meta$flywire.xyz)
+    sides = meta$side[side.indices]
+    gs$side[!is.na(side.indices)] = sides[!is.na(side.indices)]
+  }
+
   # Get FAFBv14 coordinates
-  if(length(skids)){
+  if(length(skids) & catmaid.update){
     cats = nat::neuronlist()
     batches = split(1:length(skids), round(seq(from = 1, to = 100, length.out = length(skids))))
     all.ids = c()
     for(i in 1:10){
       # Read CATMAID neurons
       message("Batch:", i, "/10")
-      cat = tryCatch(catmaid::read.neurons.catmaid(skids[batches[[i]]], OmitFailures = TRUE), error = function(e) NULL)
+      search = skids[batches[[i]]]
+      cat = tryCatch(catmaid::read.neurons.catmaid(search, OmitFailures = TRUE), error = function(e) NULL)
       if(!length(cat)){
         next
       }
@@ -39,27 +48,20 @@ flywire_matching_rewrite <- function(flywire.ids = names(flywire_neurons()),
       flywire.xyz = apply(branchpoints.flywire.raw, 1, paste_coords)
 
       # Add
-      indices = match(rownames(branchpoints),gs$skid)
-      flywire.xyz.na = is.na(gs[indices,]$flywire.xyz)
-      indices.na = indices[flywire.xyz.na]
+      indices = match(rownames(branchpoints),names(cat))
       if(length(indices)){
         gs[indices,]$FAFB.xyz = FAFB.xyz
       }
-      if(length(indices.na)){
-        gs[indices.na,]$flywire.xyz = flywire.xyz
-        gs[indices.na,]$flywire.id = fw.ids
-      }
-      if(!is.null(meta$side)){
-        side.indices = match(gs$flywire.xyz,meta$flywire.xyz)
-        sides = meta$side[side.indices]
-        gs$side[!is.na(side.indices)] = sides[!is.na(side.indices)]
+      if(length(indices)){
+        gs[indices,]$flywire.xyz = flywire.xyz
+        gs[indices,]$flywire.id = fw.ids
       }
-      all.ids=unique(c(all.ids,fw.ids))
     }
+  }
 
     # Update
-    write.cols = setdiff(colnames(gs),c("FAFB.xyz", "flywire.xyz", "flywire.id", "side"))
-    if(length(all.ids) & length(write.cols)){
+    write.cols = intersect(c("FAFB.xyz", "flywire.xyz", "flywire.id", "side"),colnames(gs))
+    if(length(write.cols)){
       for(column in write.cols){
         letter = LETTERS[match(column,colnames(gs))]
         range = paste0(letter,2,":",letter,nrow(gs)+1)
@@ -70,17 +72,41 @@ flywire_matching_rewrite <- function(flywire.ids = names(flywire_neurons()),
                             sheet = "FAFB",
                             col_names = FALSE)
       }
+      # Add flywire match to the hemibrain and LM sheets
+      ## Read the FAFB Google Sheet
+      fg = hemibrain_match_sheet(sheet = "FAFB", selected_file = selected_file)
+    }else{
+      fg = gs
     }
 
-    # Add flywire match to the hemibrain and LM sheets
-    ## Read the FAFB Google Sheet
-    fg = hemibrain_match_sheet(sheet = "FAFB", selected_file = selected_file)
-  }else{
-    all.ids = c()
-    fg = gs
+  # Figure out duplicate entries
+  fg$index = 1:nrow(fg)+1
+  dupes = unique(fg$flywire.id[duplicated(fg$flywire.id)])
+  dupes = dupes[!is.na(dupes)]
+  dupes = dupes[!dupes%in%c("NA"," ","")]
+  for(dupe in dupes){
+    sub = subset(fg, fg$flywire.id == dupe)
+    skd = unique(sub$skid)
+    skd = skd[!is.na(skd)]
+    skd = skd[!skd%in%c("NA"," ","")]
+    if(length(skd)>1){
+      next
+    }
+    best = which.max(apply(sub, 1, function(r) sum(is.na(r[c("hemibrain.match", "hemibrain.match.quality",
+                                                   "LM.match", "LM.match.quality",
+                                                   "FAFB.hemisphere.match", "FAFB.hemisphere.match.quality")]))))
+    remove = sub[-best,]
+    for(r in remove$index){
+      range.del = googlesheets4::cell_rows(r)
+      googlesheets4::range_delete(ss = selected_file,
+                                  range = range.del,
+                                  sheet = "FAFB")
+      message("Removing a row for: ", dupe)
+    }
   }
 
   # Add missing flywire information
+  all.ids = unique(gs$flywire.id)
   missing = setdiff(flywire.ids, all.ids)
   hemibrain_matching_add(ids = missing, meta = meta, dataset="flywire", selected_file = selected_file, ...)
 

R/hemibrain_matching.R

@@ -1104,6 +1104,7 @@ lm_matches <- function(priority = c("hemibrain","lm"), selected_file = options()
 #' @param flycircuit.ids flycircuit IDs to add to Google sheet if not already present.
 #' @param meta meta data for the given flycircuit IDs.
 #' @param top.nblast logical. Whether or not to also give the top NBLAST match for each entry.
+#' @param catmaid.update logical. Whether or not to update \code{flywire.xyz} and \code{flywire.id} columns, based on e CATMAID neuron specified by a \code{skid} column.
 #'
 #' @param ... arguments passed to methods for, for example, \code{neuprintr::neuprint_get_meta} and \code{elmr::fafb_get_meta}.
 #'

_pkgdown.yml

@@ -101,6 +101,7 @@ reference:
     - 'flywire_neurons'
     - 'flywire_googledrive_data'
     - 'flywire_ids_update'
+    - '`flywire_basics`'
   - title: Access light-level google drive data
     desc: ~
     contents:

data-raw/information.R

@@ -158,7 +158,7 @@ usethis::use_data(hemibrain_glomeruli_summary, overwrite = TRUE)
 ### Supplementary data for Schlegel and Bates 2021:
 supp.cols = c("bodyid", "pre", "post", "upstream", "downstream",
               "status", "name",  "voxels", "soma", "side",
-              "connectivity.type", "cell.type", "anatomy.group", "class", "cellBodyFiber",
+              "connectivity.type", "cell.type", "group","anatomy.group", "class", "cellBodyFiber",
               # "ItoLee_Lineage", "ItoLee_Hemilineage", "Hartenstein_Lineage", "Hartenstein_Hemilineage",
               "putative.classic.transmitter", "putative.other.transmitter", "glomerulus", "presyn.glom",
               "FAFB.match", "FAFB.match.quality", "layer", "ct.layer",