Merge pull request #47 from aberHRML/devel

v1.0.1

Author: jasenfinch

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: assignments
 Title: Molecular Formula Assignment For High Resolution ESI-MS Based Metabolomics Data
-Version: 1.0.0
+Version: 1.0.1
 Authors@R: person("Jasen", "Finch", email = "jsf9@aber.ac.uk", role = c("aut", "cre"))
 Description: A molecular formula assignment approach for electrospray ionisation high resolution mass spectrometry based metabolomics data.
 Depends: R (>= 3.5.0),

NAMESPACE

@@ -68,6 +68,7 @@ importFrom(dplyr,full_join)
 importFrom(dplyr,group_by)
 importFrom(dplyr,group_split)
 importFrom(dplyr,inner_join)
+importFrom(dplyr,join_by)
 importFrom(dplyr,left_join)
 importFrom(dplyr,mutate)
 importFrom(dplyr,mutate_at)

NEWS.md

@@ -1,3 +1,17 @@
+# assignments 1.0.1
+
+* The default ppm threshold has been reduced to 4.
+
+* The importance of the ^37^Cl adduct has been increased in the default negative mode adducts.
+
+* The ^13^C2 isotope has been removed from the default isotopes.
+
+* The default retention time difference limit for relationships has been changed to 2 seconds for RP-LC-HRMS and NP-LC-HRMS.
+
+* The absolute values of correlation coefficients are now used to calculate average component weights.
+
+* Where components contain a feature represented by more than one adduct and isotope combination, only the node with the highest AIS is now retained.
+
 # assignments 1.0.0
 
 * Added a `NEWS.md` file to track changes to the package.

R/components.R

@@ -48,7 +48,7 @@ clean <- function(graph,adduct_rules_table){
     cleaned_graph <- cleaned_graph %>% 
       bind_graphs() 
   }
-   
+  
   return(cleaned_graph)
 }
 
@@ -60,14 +60,14 @@ nComponents <- function(graph){
 
 componentMetrics <- function(component,max_add_iso_total){
   component %>% 
-  mutate(Size = graph_size(),
-         Nodes = n(),
-         Degree = avg_degree(Nodes,Size),
-         Density = (2 * Size) / (Nodes * (Nodes - 1)),
-         Weight = sum(Weight) / Nodes,
-         AIS = sum(AIS) / max_add_iso_total,
-         `Component Plausibility` = plausibility(Degree,AIS,Weight)
-         )
+    mutate(Size = graph_size(),
+           Nodes = n(),
+           Degree = avg_degree(Nodes,Size),
+           Density = (2 * Size) / (Nodes * (Nodes - 1)),
+           Weight = sum(Weight) / Nodes,
+           AIS = sum(AIS) / max_add_iso_total,
+           `Component Plausibility` = plausibility(Degree,AIS,Weight)
+    )
 }
 
 componentFilters <- function(){
@@ -77,6 +77,38 @@ componentFilters <- function(){
          Direction = c(rep('max',2),'min'))
 }
 
+#' @importFrom dplyr join_by
+
+deduplicate <- function(graph){
+  dedup_nodes <- graph %>% 
+    activate(nodes) %>%
+    as_tibble() %>%
+    group_split(
+      Component,
+      Feature
+    ) %>% 
+    future_map_dfr(
+      ~.x %>% 
+        arrange(
+          desc(
+            AIS
+          )
+        ) %>% 
+        slice(1)    
+    )
+  
+  graph %>% 
+    activate(nodes) %>% 
+    inner_join(
+      dedup_nodes,
+      by = join_by(
+        name, Feature, RetentionTime,
+        Isotope, Adduct, MF, `Theoretical M`, `Measured M`,
+        `Theoretical m/z`, `Measured m/z`, `PPM error`, 
+        `MF Plausibility (%)`, AIS, ID, Component)
+    )
+}
+
 #' @importFrom tidygraph as_tbl_graph activate morph unmorph graph_size group_components tbl_graph
 #' @importFrom magrittr set_names
 
@@ -88,6 +120,7 @@ calcComponents <- function(graph_nodes,
     activate(nodes) %>%
     left_join(graph_nodes,by = c('name' = 'Name')) %>%
     mutate(Component = group_components()) %>% 
+    deduplicate() %>% 
     clean(adductRules(assignment))
 
   if (nComponents(graph) > 0){
@@ -102,6 +135,7 @@ calcComponents <- function(graph_nodes,
           filter(Component == .x) %>%
           edges() %>% 
           .$coefficient %>% 
+          abs() %>% 
           mean() %>%
           tibble(Weight = .)
       },graph = graph) %>%

R/parameters.R

@@ -41,12 +41,12 @@ setClass('AssignmentParameters',
                                           maxCor = Inf),
            max_M = 800,
            MF_rank_threshold = 3,
-           ppm = 6,
+           ppm = 4,
            limit = 0.001,
            RT_diff_limit = numeric(),
-           isotopes = c('13C','18O','13C2'),
-           adducts = list(n = c("[M-H]1-", "[M+Cl]1-", "[M+K-2H]1-", 
-                                "[M-2H]2-", "[M+Cl37]1-","[2M-H]1-"),
+           isotopes = c('13C','18O'),
+           adducts = list(n = c("[M-H]1-", "[M+Cl]1-", "[M+Cl37]1-", "[M+K-2H]1-", 
+                                "[M-2H]2-", "[2M-H]1-"),
                           p = c('[M+H]1+','[M+K]1+','[M+Na]1+','[M+K41]1+',
                                 '[M+2H]2+','[2M+H]1+')),
            transformations = transformation_rules()$`MF Change`,
@@ -548,12 +548,12 @@ assignmentParameters <- function(technique = availableTechniques()){
                        `FIE-HRMS` = new('AssignmentParameters'),
                        `RP-LC-HRMS` = new('AssignmentParameters',
                                           technique = 'RP-LC-HRMS',
-                                          RT_diff_limit = 1/60),
+                                          RT_diff_limit = 2/60),
                        `NP-LC-HRMS` = new('AssignmentParameters',
                                           technique = 'NP-LC-HRMS',
-                                          RT_diff_limit = 1/60,
-                                          adducts = list(n = c("[M-H]1-", "[M+Cl]1-", "[M+K-2H]1-", 
-                                                               "[M-2H]2-", "[M+Cl37]1-","[2M-H]1-"),
+                                          RT_diff_limit = 2/60,
+                                          adducts = list(n = c("[M-H]1-", "[M+Cl]1-", "[M+Cl37]1-", "[M+K-2H]1-", 
+                                                               "[M-2H]2-","[2M-H]1-"),
                                                          p = c('[M+H]1+','[M+K]1+','[M+Na]1+','[M+K41]1+',
                                                                '[M+NH4]1+','[M+2H]2+','[2M+H]1+'))))