Lassa revs 2

shared/lassa_functions.R

@@ -3,7 +3,7 @@
 data_curation <- function(articles, outbreaks, models, parameters, plotting,switch_first_surname=FALSE) {
 
   articles   <- articles %>%
-    mutate(refs = paste(first_author_first_name," (",year_publication,")",sep="")) %>% #define references
+    mutate(refs = paste(first_author_surname," (",year_publication,")",sep="")) %>% #define references
     group_by(refs) %>% mutate(counter = row_number()) %>% ungroup() %>% #distinguish same-author-same-year references
     mutate(new_refs = ifelse(refs %in% refs[duplicated(refs)], paste0(sub("\\)$", "", refs),letters[counter],")"), refs)) %>%
     select(-counter,-refs) %>% rename(refs = new_refs)

src/db_compilation/lassa_cleaning.R

@@ -89,20 +89,20 @@ lassa_cleaning <- function(df) {
       TRUE ~ paper_copy_only
     )) %>%
     # name typos
-    mutate(first_author_surname = case_when(
+    mutate(first_author_first_name = case_when(
       covidence_id %in% 2648 ~ "Shirley C.",
       covidence_id %in% 1447 ~ "N.A.",
-      TRUE ~ first_author_surname
+      TRUE ~ first_author_first_name
     )) %>%
-    mutate(first_author_first_name = case_when(
+    mutate(first_author_surname = case_when(
       covidence_id %in% 2648 ~ "Nimo-Paintsil",
       covidence_id %in% 2585 ~ "Dalhat",
       covidence_id %in% 1033 ~ "Ehichioya",
       covidence_id %in% 661 ~ "Kerneis",
-      TRUE ~ first_author_first_name
+      TRUE ~ first_author_surname
     )) %>%
-    mutate(first_author_first_name = sub(".*\\.(.*)", "\\1", first_author_first_name)) %>%
-    mutate(first_author_first_name = sub("^\\s+", "", first_author_first_name)) 
+    mutate(first_author_surname = sub(".*\\.(.*)", "\\1", first_author_surname)) %>%
+    mutate(first_author_surname = sub("^\\s+", "", first_author_surname)) 
      # revised qa after parameter removed: now outbreak only
   df[df$covidence_id %in% 152, c("qa_m1", "qa_m2", "qa_a3", "qa_a4", "qa_d6", "qa_d7")] <- NA
   df

src/db_compilation/lassa_qa_fixing.csv

@@ -1,4 +1,4 @@
-Article_ID,Pathogen,Covidence_ID,FirstAauthor_Surname,Article_title,DOI,Journal,Year_publication,Volume,Issue,Page_first,Page_last,Name_data_entry,Paper_copy_only,Notes,FirstAuthor_FirstName,QA_M1,QA_M2,QA_A3,QA_A4,QA_D5,QA_D6,QA_D7,ID,Covidence_ID_text,fixed,num_rows,matching
+Article_ID,Pathogen,Covidence_ID,FirstAuthor_FirstName,Article_title,DOI,Journal,Year_publication,Volume,Issue,Page_first,Page_last,Name_data_entry,Paper_copy_only,Notes,FirstAauthor_Surname,QA_M1,QA_M2,QA_A3,QA_A4,QA_D5,QA_D6,QA_D7,ID,Covidence_ID_text,fixed,num_rows,matching
 1,Lassa mammarenavirus,18,PA,Lassa Fever in Children in Sierra-Leone; West-Africa,NA,Transactions of the Royal Society of Tropical Medicine and Hygiene,1986,80,4,577,582,Christian,FALSE,"Study in Children
 Seasonal cluster in April; May and August",Webb,Yes,No,NA,NA,Yes,No,No,9f62e47ed45a53c1336e8e63d7d7f6cb,18,1,,
 4,Lassa mammarenavirus,49,Fransje,Hemorrhagic-Fever Virus-Infections in an Isolated Rain-Forest Area of Central Liberia - Limitations of the Indirect Immunofluorescence Slide Test for Antibody Screening in Africa,NA,Tropical and Geographical Medicine,1986,38,1,209,214,Christian,TRUE,"Link with Epilepsy

src/db_compilation/sars_cleaning.R

@@ -105,18 +105,18 @@ sars_cleaning <- function(df) {
         #   covidence_id %in% c(845,917) ~ FALSE,
         #   TRUE ~ paper_copy_only)) %>%
         #name typos
-        mutate(first_author_surname = case_when(
+        mutate(first_author_first_name = case_when(
         #   covidence_id == 2648 ~ 'Shirley C.',
            covidence_id == 6909 ~ 'MEC', #The Chinese SARS Molecular Epidemiology Consortium
-           TRUE ~ first_author_surname)) %>%
-        mutate(first_author_first_name = sub(".*\\.(.*)", "\\1", first_author_first_name, useBytes = TRUE)) %>%
-        mutate(first_author_first_name = sub("^\\s+", "", first_author_first_name, useBytes = TRUE)) %>%                 
-       mutate(first_author_first_name = case_when(
+           TRUE ~ first_author_first_name)) %>%
+        mutate(first_author_surname = sub(".*\\.(.*)", "\\1", first_author_surname, useBytes = TRUE)) %>%
+        mutate(first_author_surname = sub("^\\s+", "", first_author_surname, useBytes = TRUE)) %>%                 
+       mutate(first_author_surname = case_when(
       #   covidence_id == 2648 ~ 'Nimo-Paintsil',
       #   covidence_id == 2585 ~ 'Dalhat',
       #   covidence_id == 1033 ~ 'Ehichioya',
          covidence_id == 6909 ~ 'MEC', #The Chinese SARS Molecular Epidemiology Consortium
-         TRUE ~ first_author_first_name))     
+         TRUE ~ first_author_surname))     
     df
 }
 

src/db_extraction/orderly.R

@@ -325,6 +325,11 @@ if (pathogen == "EBOLA") {
   models$Covidence_ID[   models$Covidence_ID    == 3158] <- 3153
   params$Covidence_ID[   params$Covidence_ID    == 3158] <- 3153
 
+  #correct author first/last names
+  articles <- articles %>% rename(temp_col = FirstAuthor_FirstName,
+                                  FirstAuthor_FirstName = FirstAauthor_Surname) %>%
+                           rename(FirstAauthor_Surname = temp_col)
+
   #blank outbreaks
   outbreaks <- outbreaks %>% filter(!(Covidence_ID == 845 & Outbreak_ID == 1))
 
@@ -343,7 +348,15 @@ if (pathogen == "EBOLA") {
 
   #missing parameter types
   params$Parameter_type[params$Covidence_ID==2684&params$access_param_id==37] <- 'Risk factors'
-}  
+} else if (pathogen == "SARS") {
+
+  #correct author first/last names
+  articles <- articles %>% rename(temp_col = FirstAuthor_FirstName,
+                                  FirstAuthor_FirstName = FirstAauthor_Surname) %>%
+                           rename(FirstAauthor_Surname = temp_col)  
+
+}
+
 # Pathogen-specific cleaning
 articles <- clean_articles(articles)
 models <- clean_models(models)

src/lassa_collate/orderly.R

@@ -6,12 +6,12 @@ orderly_parameters(pathogen = NULL)
 orderly_dependency("lassa_serology", "latest(parameter:pathogen == this:pathogen)",
                    files = c("figure_2.png","figure_2.pdf"))
 orderly_dependency("lassa_severity", "latest(parameter:pathogen == this:pathogen)",
-                   files = c("figure_3.png","figure_S6.png","figure_S7.png","figure_S8.png",
-                             "figure_S9.png","figure_S10.png","figure_S11.png","figure_S12.png",
-                             "figure_3.pdf","figure_S6.pdf","figure_S7.pdf","figure_S8.pdf",
-                             "figure_S9.pdf","figure_S10.pdf","figure_S11.pdf","figure_S12.pdf"))
+                   files = c("figure_3.png","figure_S7.png","figure_S8.png",
+                             "figure_S9.png","figure_S10.png","figure_S11.png","figure_S12.png","figure_S13.png","figure_S14.png",
+                             "figure_3.pdf","figure_S7.pdf","figure_S8.pdf",
+                             "figure_S9.pdf","figure_S10.pdf","figure_S11.pdf","figure_S12.pdf","figure_S13.pdf","figure_S14.pdf"))
 orderly_dependency("lassa_delays", "latest(parameter:pathogen == this:pathogen)",
-                   files = c("figure_4.png","figure_S5.png","figure_4.pdf","figure_S5.pdf"))
+                   files = c("figure_4.png","figure_S5.png","figure_S6.png","figure_4.pdf","figure_S5.pdf","figure_S6.pdf"))
 orderly_dependency("lassa_transmission", "latest(parameter:pathogen == this:pathogen)",
                    files = c("figure_5.png","figure_5.pdf"))
 orderly_dependency("lassa_summary", "latest(parameter:pathogen == this:pathogen)",
@@ -26,10 +26,12 @@ orderly_artefact("lassa-specific outputs",
                    "main/figure_2.pdf","main/figure_3.pdf","main/figure_4.pdf","main/figure_5.pdf",
                    "supp/figure_S2.png","supp/figure_S3.png","supp/figure_S4.png",
                    "supp/figure_S5.png","supp/figure_S6.png","supp/figure_S7.png","supp/figure_S8.png",
-                   "supp/figure_S9.png","supp/figure_S10.png","supp/figure_S11.png","supp/figure_S12.png",
+                   "supp/figure_S9.png","supp/figure_S10.png","supp/figure_S11.png","supp/figure_S12.png","supp/figure_S13.png",
+                   "supp/figure_S14.png",
                    "supp/figure_S2.pdf","supp/figure_S3.pdf","supp/figure_S4.pdf",
                    "supp/figure_S5.pdf","supp/figure_S6.pdf","supp/figure_S7.pdf","supp/figure_S8.pdf",
-                   "supp/figure_S9.pdf","supp/figure_S10.pdf","supp/figure_S11.pdf","supp/figure_S12.pdf",
+                   "supp/figure_S9.pdf","supp/figure_S10.pdf","supp/figure_S11.pdf","supp/figure_S12.pdf","supp/figure_S13.pdf",
+                   "supp/figure_S14.pdf",
                    "supp/latex_outbreaks.csv","supp/latex_models.csv",
                    "supp/latex_transmission.csv","supp/latex_delays.csv","supp/latex_severity.csv",
                    "supp/latex_seroprevalence.csv","supp/latex_riskfactors.csv"))
@@ -86,6 +88,10 @@ file.copy(from = "figure_S11.png",
           to = "supp/figure_S11.png")
 file.copy(from = "figure_S12.png",
           to = "supp/figure_S12.png")
+file.copy(from = "figure_S13.png",
+          to = "supp/figure_S13.png")
+file.copy(from = "figure_S14.png",
+          to = "supp/figure_S14.png")
 
 file.copy(from = "figure_S2.pdf",
           to = "supp/figure_S2.pdf")
@@ -109,6 +115,10 @@ file.copy(from = "figure_S11.pdf",
           to = "supp/figure_S11.pdf")
 file.copy(from = "figure_S12.pdf",
           to = "supp/figure_S12.pdf")
+file.copy(from = "figure_S13.pdf",
+          to = "supp/figure_S13.pdf")
+file.copy(from = "figure_S14.pdf",
+          to = "supp/figure_S14.pdf")
 
 file.copy(from = "latex_outbreaks.csv",
           to = "supp/latex_outbreaks.csv")

src/lassa_delays/orderly.R

@@ -103,11 +103,24 @@ patchwork <- patchwork + plot_annotation(tag_levels = 'A')
 ggsave("figure_4.png", plot = patchwork, width = 12, height = 16)
 ggsave("figure_4.pdf", plot = patchwork, width = 12, height = 16)
 
-#figure_S5: hypothetical probability distributions for onset-admission delay
+# #figure_S5: hypothetical probability distributions for onset-admission delay
 p1 <- pdf_generic(m5$result,"common",c("Gamma","Lognormal","Weibull"),c(-0.2,30),'Onset-Admission Delay (days)')
 p2 <- pdf_generic(m5$result,"random",c("Gamma","Lognormal","Weibull"),c(-0.2,30),'Onset-Admission Delay (days)')
 
 patchwork <- (p1 + p2) + plot_layout(ncol = 2, widths = c(1,1))
 patchwork <- patchwork + plot_annotation(tag_levels = 'A')
 ggsave("figure_S5.png", plot = patchwork, width = 12, height = 6)
-ggsave("figure_S5.pdf", plot = patchwork, width = 12, height = 6)
+ggsave("figure_S5.pdf", plot = patchwork, width = 12, height = 6)
+
+#figure_S6: meta-analysis funnel plot
+png(file = "temp.png", width = 9500, height = 6500, res = 1000)
+funnel(m5$result, xlim = c(5,12), ylim = c(0,1.5), common = FALSE,#only plots funnel for either common or random for some reason
+       pch = 22, bg = "dodgerblue3", level = 0.95, 
+       studlab = TRUE, cex.studlab = 0.75, pos.studlab = 2)
+dev.off()
+gg <- png::readPNG("temp.png", native = TRUE)
+file.remove("temp.png")
+gg <- wrap_elements(plot = rasterGrob(gg, interpolate = TRUE))
+
+ggsave("figure_S6.png", plot = gg, width = 12, height = 8)
+ggsave("figure_S6.pdf", plot = gg, width = 12, height = 8)

src/lassa_serology/orderly.R

@@ -110,19 +110,35 @@ n1 <- data.frame(latitude  = c(19.1,18.1,17.1,15.1,15.0,12.5,
                                "Central\nAfrican\nRepublic","Gabon","Equatorial\nGuinea","Cameroon","Côte d'Ivoire","Ghana",
                                "Republic\nof Congo","Benin","Togo","Nigeria","Guinea","Gambia",
                                "Guinea\n-Bissau","Sierra\nLeone","Liberia"))
-n2 <- data.frame(latitude = numeric(0), longitude = numeric(0), c_name = character(0))
+n2 <- data.frame(latitude  = c(16.4,
+                               3.8,4.25,
+                               5.2,3.8,9.0,
+                               7.0,5), 
+                 longitude = c(-11.5,
+                               -4.5,-0.5,
+                               2.3,4.5,-15.5,
+                               -14.5,-11.5), 
+                 c_name    = c("Mali",
+                               "Côte d'Ivoire","Ghana",
+                               "Benin","Nigeria","Guinea",
+                               "Sierra\nLeone","Liberia"))
 
 #specify position of lines pointing to countries (added to map after function call)
-ll <- data.frame(lat1 = c(2.50,-3.2,1.0,5.1,3.3,13.4,11.500,7.200,4.5),
-                 lon1 = c(23.5,7.30,7.6,1.6,3.7,-17.8,-17.7,-14.5,-11),
-                 lat2 = c(3.80,-2.5,1.7,6.0,4.5,13.4,11.700,7.900,5.5),
-                 lon2 = c(22.5,9.20,9.4,1.4,5.3,-17,-16.500,-13.2,-10))
+ll1 <- data.frame(lat1 = c(2.50,-3.2,1.0,5.1,3.3,13.4,11.500,7.200,4.5),
+                  lon1 = c(23.5,7.30,7.6,1.6,3.7,-17.8,-17.7,-14.5,-11),
+                  lat2 = c(3.80,-2.5,1.7,6.0,4.5,13.4,11.700,7.900,5.5),
+                  lon2 = c(22.5,9.20,9.4,1.4,5.3,-17,-16.500,-13.2,-10))
+ll2 <- data.frame(lat1 = c( 16.1,  9.1,   7.45, 5.1, 4.1, 4.5, 5.4, 4),
+                  lon1 = c(-11.4,-14.75,-13.9,-10.8,-4.5,-0.5, 2.3, 4.5),
+                  lat2 = c( 15.3,  9.7,   7.9,  5.7, 5,   5.1, 6.2, 4.8),
+                  lon2 = c(-10.9,-14.0, -13.2,-10.2,-4.9,-0.8, 2.2, 5.3))
 
 #call mapping function
 p1 <- map_generic(l0,l1,d1,f1,n1,range_mp=TRUE,summ_dups="mean",c(-23.2,25.1),c(-3.9,22.25),c(0,60),'magma','')
 p1 <- p1 + geom_rect(aes(xmin = -18.8, xmax = 14.75, ymin = 2, ymax = 16.75), linetype = 2, color = "grey40", fill = NA)
-p1 <- p1 + geom_segment(data = ll, aes(x = lon1, xend = lon2, y = lat1, yend = lat2), size = 0.5, color = "black")
+p1 <- p1 + geom_segment(data = ll1, aes(x = lon1, xend = lon2, y = lat1, yend = lat2), size = 0.5, color = "black")
 p2 <- map_generic(l0,l1,d2,f2,n2,range_mp=TRUE,summ_dups="mean",c(-18.5,13.13),c(2.75,16.1),c(0,60),'magma','')
+p2 <- p2 + geom_segment(data = ll2, aes(x = lon1, xend = lon2, y = lat1, yend = lat2), size = 0.5, color = "black")
 
 patchwork <- p1 / p2
 patchwork <- patchwork + plot_layout(heights = c(1,0.75)) + plot_annotation(tag_levels = 'A')

src/lassa_severity/orderly.R

@@ -28,12 +28,12 @@ orderly_shared_resource("Africa_Boundaries-shp" = "Africa_Boundaries-shp",
                         "rivers_africa_37333"   = "rivers_africa_37333")
 orderly_shared_resource("lassa_functions.R" = "lassa_functions.R")
 source("lassa_functions.R")
-orderly_artefact("lassa-specific figures",c("figure_3.png","figure_S6.png","figure_S7.png",
+orderly_artefact("lassa-specific figures",c("figure_3.png","figure_S7.png",
                                             "figure_S8.png","figure_S9.png","figure_S10.png",
-                                            "figure_S11.png","figure_S12.png",
-                                            "figure_3.pdf","figure_S6.pdf","figure_S7.pdf",
+                                            "figure_S11.png","figure_S12.png","figure_S13.png","figure_S14.png",
+                                            "figure_3.pdf","figure_S7.pdf",
                                             "figure_S8.pdf","figure_S9.pdf","figure_S10.pdf",
-                                            "figure_S11.pdf","figure_S12.pdf"))
+                                            "figure_S11.pdf","figure_S12.pdf","figure_S13.pdf","figure_S14.pdf"))
 
 ###################
 ## DATA CURATION ##
@@ -183,7 +183,7 @@ patchwork <- patchwork + plot_annotation(tag_levels = 'A')
 ggsave("figure_3.png", plot = patchwork, width = 18, height = 12)
 ggsave("figure_3.pdf", plot = patchwork, width = 18, height = 12)
 
-#figure_S6-S10: meta-analysis with all estimates plotted
+#figure_S7-S11: meta-analysis with all estimates plotted
 m1 <- metaprop_wrap(dataframe = da %>% arrange(-central), subgroup = "lineage", 
                     plot_pooled = TRUE, sort_by_subg = TRUE, plot_study = TRUE, digits = 3, colour = "red", 
                     width = 1200, height = 1600, resolution = 115)
@@ -206,18 +206,18 @@ p3 <- m3$plot
 p4 <- m4$plot
 p5 <- m5$plot
 
-ggsave("figure_S6.png", plot = p1, width = 12, height = 16)
-ggsave("figure_S6.pdf", plot = p1, width = 12, height = 16)
-ggsave("figure_S7.png", plot = p2, width = 12, height = 16)
-ggsave("figure_S7.pdf", plot = p2, width = 12, height = 16)
-ggsave("figure_S8.png", plot = p3, width = 12, height = 16)
-ggsave("figure_S8.pdf", plot = p3, width = 12, height = 16)
-ggsave("figure_S9.png", plot = p4, width = 12, height = 16)
-ggsave("figure_S9.pdf", plot = p4, width = 12, height = 16)
-ggsave("figure_S10.png", plot = p5, width = 12, height = 16)
-ggsave("figure_S10.pdf", plot = p5, width = 12, height = 16)
+ggsave("figure_S7.png", plot = p1, width = 12, height = 16)
+ggsave("figure_S7.pdf", plot = p1, width = 12, height = 16)
+ggsave("figure_S8.png", plot = p2, width = 12, height = 16)
+ggsave("figure_S8.pdf", plot = p2, width = 12, height = 16)
+ggsave("figure_S9.png", plot = p3, width = 12, height = 16)
+ggsave("figure_S9.pdf", plot = p3, width = 12, height = 16)
+ggsave("figure_S10.png", plot = p4, width = 12, height = 16)
+ggsave("figure_S10.pdf", plot = p4, width = 12, height = 16)
+ggsave("figure_S11.png", plot = p5, width = 12, height = 16)
+ggsave("figure_S11.pdf", plot = p5, width = 12, height = 16)
 
-#figure_S11: meta-analysis with only known duplicates excluded
+#figure_S12: meta-analysis with only known duplicates excluded
 db <- d1 %>% filter(duplicate_cfr %in% c("False","Assumed"))
 
 m1 <- metaprop_wrap(dataframe = db, subgroup = "lineage", 
@@ -240,10 +240,10 @@ p4 <- m4$plot
 
 patchwork <- (p1 + p2 + p3 + p4) + plot_layout(ncol = 2, widths = c(1,1))
 patchwork <- patchwork + plot_annotation(tag_levels = 'A') 
-ggsave("figure_S11.png", plot = patchwork, width = 12, height = 12)
-ggsave("figure_S11.pdf", plot = patchwork, width = 12, height = 12)
+ggsave("figure_S12.png", plot = patchwork, width = 12, height = 12)
+ggsave("figure_S12.pdf", plot = patchwork, width = 12, height = 12)
 
-#figure_S12: meta-analysis without de-duplication
+#figure_S13: meta-analysis without de-duplication
 dc <- d1
 
 m1 <- metaprop_wrap(dataframe = dc, subgroup = "lineage", 
@@ -266,5 +266,18 @@ p4 <- m4$plot
 
 patchwork <- (p1 + p2 + p3 + p4) + plot_layout(ncol = 2, widths = c(1,1))
 patchwork <- patchwork + plot_annotation(tag_levels = 'A') 
-ggsave("figure_S12.png", plot = patchwork, width = 12, height = 12)
-ggsave("figure_S12.pdf", plot = patchwork, width = 12, height = 12)
+ggsave("figure_S13.png", plot = patchwork, width = 12, height = 12)
+ggsave("figure_S13.pdf", plot = patchwork, width = 12, height = 12)
+
+#figure_S14: meta-analysis funnel plot
+png(file = "temp.png", width = 9500, height = 6500, res = 1000)
+funnel(m5$result, xlim = c(-3,2), ylim = c(0,1), common = FALSE,#only plots funnel for either common or random for some reason
+       pch = 22, bg = "red", level = 0.95, 
+       studlab = TRUE, cex.studlab = 0.75, pos.studlab = 2)
+dev.off()
+gg <- png::readPNG("temp.png", native = TRUE)
+file.remove("temp.png")
+gg <- wrap_elements(plot = rasterGrob(gg, interpolate = TRUE))
+
+ggsave("figure_S14.png", plot = gg, width = 12, height = 8)
+ggsave("figure_S14.pdf", plot = gg, width = 12, height = 8)

src/sars_world_map/orderly.R

@@ -36,18 +36,12 @@ articles   <- read_csv("articles.csv")
 models     <- read_csv("models.csv")
 parameters <- read_csv("parameters.csv")
 
-# to save down for epireview use plotting = FALSE
-dfs <- data_curation(articles,tibble(),models,parameters, plotting =  FALSE, switch_first_surname = TRUE )
-
-articles   <- dfs$articles
-models     <- dfs$models
-parameters <- dfs$parameters
-
 write_csv(articles,'sars_articles.csv')
 write_csv(models,'sars_models.csv')
 write_csv(parameters,'sars_parameters.csv')
 
-dfs <- data_curation(articles,tibble(),models,parameters, plotting =  TRUE, switch_first_surname = TRUE )
+# to save down for epireview use plotting = FALSE
+dfs <- data_curation(articles,tibble(),models,parameters, plotting =  FALSE)
 
 articles   <- dfs$articles
 models     <- dfs$models