adding table 3

Author: Liam Shaw

figures/Table-3-summary-best-fit-GAMs.csv

@@ -0,0 +1,40 @@
+"","","Value","Z statistic","Chi-sq statistic","P-value","Effective Degrees of Freedom","Percent Dev. Explained","model"
+"1","Intercept","1.28","23.24",NA,"<0.001",NA,NA,"Bacterial richness per species"
+"2","Intercept","1.28","23.24",NA,"<0.001",NA,NA,NA
+"3","hDiseaseZACitesLn",NA,NA,"584.96","<0.001","3.63","81.55%",NA
+"4","hOrderCHIROPTERA",NA,NA,"7.5","0.003","0.86","1.56%",NA
+"5","hOrderDIDELPHIMORPHIA",NA,NA,"2.81","0.058","0.78","0.67%",NA
+"6","hOrderDIPROTODONTIA",NA,NA,"7.03","0.005","0.89","1.69%",NA
+"7","hOrderLAGOMORPHA",NA,NA,"1.19","0.174","0.61","0.32%",NA
+"8","hOrderPRIMATES",NA,NA,"10.36","0.002","0.97","1.8%",NA
+"9","hOrderRODENTIA",NA,NA,"47.45","<0.001","1","7.88%",NA
+"10","LnAreaHost",NA,NA,"10.89","0.117","4.13","1.88%",NA
+"11","S20",NA,NA,"46.42","<0.001","1.15","2.65%",NA
+"12","Intercept","1.11","28.65",NA,"<0.001",NA,NA,"Viral richness per species"
+"13","Intercept","1.11","28.65",NA,"<0.001",NA,NA,NA
+"14","hDiseaseZACitesLn",NA,NA,"2884.3","<0.001","4.51","76.59%",NA
+"15","hMassGramsPVR",NA,NA,"5.73","0.193","3.2","0.8%",NA
+"16","hOrderCETARTIODACTYLA",NA,NA,"35.02","<0.001","0.95","2.93%",NA
+"17","hOrderCHIROPTERA",NA,NA,"18.81","<0.001","0.97","2.05%",NA
+"18","hOrderCINGULATA",NA,NA,"1.1","0.144","0.51","0.18%",NA
+"19","hOrderDIDELPHIMORPHIA",NA,NA,"1.88","0.097","0.67","0.3%",NA
+"20","hOrderEULIPOTYPHLA",NA,NA,"8.81","0.002","0.9","1.3%",NA
+"21","hOrderPRIMATES",NA,NA,"105.73","<0.001","0.99","7.52%",NA
+"22","LnAreaHost",NA,NA,"119.01","<0.001","3.16","5.72%",NA
+"23","S20",NA,NA,"41.31","<0.001","2.96","2.61%",NA
+"24","Intercept","-1.46","-10.63",NA,"<0.001",NA,NA,"Zoonotic potential (bacteria)"
+"25","Intercept","-1.46","-10.63",NA,"<0.001",NA,NA,NA
+"26","gramGrampositive",NA,NA,"0.54","0.277","0.29","0.15%",NA
+"27","motility",NA,NA,"2.64","0.152","0.66","0.55%",NA
+"28","oxygenFacultativelyanaerobic",NA,NA,"12.04","0.01","0.9","1.79%",NA
+"29","oxygenMicroaerophilic",NA,NA,"17.75","0.005","0.92","2.04%",NA
+"30","PHB.median",NA,NA,"159.77","<0.001","4.82","47.81%",NA
+"31","PubMed.records.log",NA,NA,"151.9","<0.001","2.59","42.04%",NA
+"32","vector.borne",NA,NA,"50.49","<0.001","0.95","5.63%",NA
+"33","Intercept","-1.76","-10.59",NA,"<0.001",NA,NA,"Zoonotic potential (viruses)"
+"34","Intercept","-1.76","-10.59",NA,"<0.001",NA,NA,NA
+"35","genome.typenegativessRNA",NA,NA,"28.51","<0.001","0.99","11.28%",NA
+"36","genome.typepositivessRNA",NA,NA,"39.61","0.003","0.89","5.14%",NA
+"37","PHB.mean",NA,NA,"68.13","<0.001","3.36","36.66%",NA
+"38","SRA.records.log",NA,NA,"54.72","<0.001","3.65","33.33%",NA
+"39","vector.borne",NA,NA,"41.05","<0.001","0.95","13.59%",NA

scripts/08-tables-GAMs.R

@@ -0,0 +1,59 @@
+library(mgcv)
+library(stringi)
+library(dplyr)
+library(purrr)
+library(officer)
+source('scripts/Olival_fit_gam.R')
+
+top_models <- lapply(c("intermediates/all_bacteria_models.rds", "intermediates//all_viruses_models.rds", "intermediates/zoonotic-GAM-fits-bacteria.rds", "intermediates/zoonotic-GAM-fits-virus.rds"), function(mods) {
+  readRDS(mods)$model[[1]]
+})
+
+model_names = c("Bacterial richness per species",
+                "Viral richness per species",
+                "Zoonotic potential (bacteria)",
+                "Zoonotic potential (viruses)")
+
+model_tables = map2(top_models, model_names, function(modd, model_name) {
+  summ = summary(modd)
+  summ$p.table
+  summ$s.table
+  rel_dev = get_relative_contribs(modd)
+  
+  bind_rows(data_frame(Term = stri_extract_first_regex(rownames(summ$p.table), "(?<=\\()[^\\)]+(?=\\))"),
+                       Value = round(summ$p.table[,1], 2),
+                       `Z statistic` = round(summ$p.table[,3], 2),
+                       `Chi-sq statistic` = NA,
+                       `P-value` = ifelse(summ$p.table[,4] > 0.001, as.character(round(summ$p.table[,4], digits=3)), "<0.001"),
+                       `Effective Degrees of Freedom` = NA,
+                       `Percent Dev. Explained` = as.character(NA),
+                       model = model_name),
+            data_frame(Term = stri_extract_first_regex(rownames(summ$s.table), "(?<=s\\()[^\\)]+(?=\\))"),
+                       Value = NA,
+                       `Z statistic` = NA,
+                       `Chi-sq statistic` = round(summ$s.table[,3], 2),
+                       `P-value` = ifelse(summ$s.table[,4] > 0.001, as.character(round(summ$s.table[,4], digits=3)), "<0.001"),
+                       `Effective Degrees of Freedom` = round(summ$s.table[,1], 2),
+                       `Percent Dev. Explained`= paste0(round(100*rel_dev$rel_deviance_explained, 2), "%"),
+                       model=model_name))
+  
+})
+
+model_rows = map_int(model_tables, nrow)
+model_tables2 = model_tables %>%
+  map(~ rbind(.[1,], .)) %>%
+  map(function(x) {
+    x$model = c(x$model[1], rep(NA, nrow(x) -1))
+    return(x)
+  }) %>%
+  bind_rows %>%
+  mutate_each(funs(as.character), -Term, -model) #%>%
+  #arrange(model, Term !="Intercept") %>%
+  dplyr::select(8, 1:7)
+
+names(model_tables2)[1] <- ""
+
+print(kable(model_tables2))
+write.csv(model_tables2, file='figures/Table-3-summary-best-fit-GAMs.csv')
+
+