Coronavirus Clinical Trials R library

Description

This R project containt some script to format the database of clinical trials on covid-19 in Gargantext readable files (see http://Gargarntext.org).

The database should be in the tsv format (separator = tabulation ; no delimiters) and be formated in UTF8.

Load of Data

First define what is the name of the file to be processed. This file should be in the folder /data

library(lubridate)
source("coronalib.R") # R libraries 
name<-"Database 210120_cleaned" # name of the csv to be loaded

AllData <-read.csv(paste("data/",name,".csv",sep=""),head=TRUE,sep="\t")
AllData$Inclusion.criteria <- NULL
AllData$Exclusion.criteria <- NULL
#AllData <- filter(AllData,!is.na(AllData$Registration.date))
nrow(AllData)

[1] 8836

x <- unique(AllData$Trial.registration.number)
#head(x)[[1]]
head(AllData)

NA

Data segmentation

Several dataframe are generated according to which are the CTs under study.

library(dplyr) 
library(stringr) 
Prevention<- filter(AllData,grepl("Prevention",AllData$Study.aim)) # CTs taggés prevention
Treatments <- filter(AllData,grepl("Treatment",AllData$Study.aim)) # # CTs taggés Treaments
Posttreatment<- filter(AllData,grepl("Post treatment",AllData$Study.aim)) # CTs taggé Post-Treatment
print(paste(count(Prevention)," Prevention arms,", count(Treatments)," Treatments arms and ",count(Posttreatment)," Post-treatment arms."))

[1] "1244  Prevention arms, 4551  Treatments arms and  0  Post-treatment arms."

Export of data and viz

Data are exported in several formats. The list of all treatments is also exported assuming that treatments are separated by a ‘+’ signe in the column treatment of the original db.

source("coronalib.R")
library(reshape)
library(wordcloud)
library(ggplot2)

## Exporting for phylo
garg_export_for_phylo(AllData,"AllData")
garg_export_for_phylo_types_only(AllData,"AllData")

# Html format
# export of a corpus with treatments and outcomes
garg_export_with_html(Treatments,"Treatment") # exporte Treatmeant et Outcomes des essais cliniques de type Treatment
garg_export_with_html(AllData,"AllData") # exporte Treatmeant et Outcomes des essais cliniques de tous types
garg_export_treatments_with_html(AllData,"AllData") # exporte Treatmeant des essais cliniques de tous types


# export of the list of all types of treatments whatever the phase in the format Gargantext map list Gargantext V3 & V4
gargV4_export_treaments_list(AllData,"AllDb")
gargV3_export_treaments_list(AllData,"AllDb")
gargV3_export_treamentsTypes_list(AllData,"AllDb")


# Conversion of the tsv file into Gargantext readable tsv dile
# Seleciton of the kind of CT to export : All / Prevention / Treatment / Post-treatment
# Selection of the kind of informations to include in the main text to be processed by Gargantext (bastract column): Treatmeant and/or Outcomes


# simple txt export
garg_export_all_plain(Treatments,"Treatment") # export main information in plain text
garg_export_OnlyTreatments(Treatments,"Treatment") # export only treatments in plain text
garg_export_OnlyOutcomes(Treatments,"Treatment") # export only outcomes in plain text
garg_export_all_plain(AllData,"All") # export main information in plain text

# raw export (just to have specific maps)
garg_export_raw_treatments(AllData,"AllData") ## export only info relative to treatments without any formating.

# Some simple viz - Tag cloud of the treaments per category of CT
TreatmentsCloud(Treatments)

TreatmentsCloud(Prevention)

TreatmentsCloud(Posttreatment)

Factor `treatments` contains implicit NA, consider using `forcats::fct_explicit_na`Error in strwidth(words[i], cex = size[i], ...) : valeur 'cex' incorrecte