- Reference +
- + + Articles + + + +
- Changelog diff --git a/main/articles/index.html b/main/articles/index.html new file mode 100644 index 00000000..b6f287f4 --- /dev/null +++ b/main/articles/index.html @@ -0,0 +1,80 @@ + +
- Visit autoslideR-demo. +Clone the project and open it on OCEAN R 4.3.1. +
- Run
+
renv::restore(repos = c(CRAN = "http://cran.rstudio.com", RSPM = "https://rspm.roche.com/Non-Validated/latest"))+from you console to construct the environment, e.g. installing the +required packages.
+
+ - Modify the files in the folder
IMmotion010then run +run_slide.R. The folder contains the following +objects:
+ -
+
atezo.pptx: The slide theme template. You can replace +this file by your own template. The slide master of the theme template +should contains the basic layouts: “Title and Content” and “Two +Content”. For further details, refer to the PD +Connect post How to create a PowerPoint theme for autoslideR +deck. +
+ -
+
filters.yml: To define and store filters.
+ -
+
spec.yml: To define the content (TLG) on each slide. +Each program corresponds to a TLG. The order of the program in +spec.ymldetermines the order of the contents in the final +slide deck. User can also specify the title, footnote, filter (using +suffix), and argument values to be passed to the program.
+ -
+
metadata.yml: Study related metadata information, +including entimice data paths, lopo paths, TA and study number, +etc.
+ -
+
programs: This folder hosts your R scripts for the +automatic slide generation. We recommend to have two separate workflows: +one for data pre-processing, the other for executing +spec.ymlto generate slides.
+ -
+
adsl: ADSL dataset, dataframe.
+
+ -
+
arm: Arm variable, character."TRT01P"by +default.
+
+ -
+
vars: A vector of characters, names of variables to be +summarized in the table.
+
+ -
+
stats: A vector of characters, statistics to be +calculated for the variables in argumentvars. Common ones +are"n","median","mean_sd", +"range","count_fraction". The default is +c("median", "range", "count_fraction"). Please see +.statsfromtern::analyze_vars()+for more details.
+ -
+
adsldataset has following variables:
+
+ -
+
USUBJID: Character.
+
+ -
+
STUDYID: Character.
+
+ - The variables in arguments
armandvars+exist inadsldataset.
+
+ - In the targeted population, the
armvariable in the +adslshould have no NAs or no strings with whitespaces +only.
+
+ - Although not required, a good practice is to convert all character
+variables to use into factors.
+
+ -
+
adsl: ADSL dataset, dataframe.
+
+ -
+
arm: Arm variable, character."TRT01A"by +default.
+
+ -
+
colcount: Whether to add col counts, logical. +"FALSE"by default.
+
+ -
+
drug_names: A character vector, name of the drugs to be +displayed in the header, e.g.: +c("Drug X", "Placebo", "Combination").
+
+ -
+
drug_vars: A character vector, corresponding values of +each drug in thearmvariable, e.g.: +c("A: Drug X", "B: Placebo", "C: Combination").
+
+ -
+
drug_sdt: A character vector, names of variables that +represent the start day of each drug, which can be different for +different drugs.
+
+ -
+
drug_discfl: A character vector, names of variables +that represent the treatment discontinuation flag of each drug. Each +variable has only"Y"/"N".
+
+ -
+
drug_discst: A character vector, names of variables +that represent the treatment discontinuation status of each drug. +Example values are"COMPLETED", +"DISCONTINUED","ONGOING".
+
+ -
+
drug_discrs: A character vector, names of variables +that represent the treatment discontinuation reason of each drug.
+ The variables in arguments
arm, +drug_vars,drug_sdt,drug_discfl, +drug_discst,drug_discrsexist in +adsldataset.
+The length of
drug_vars,drug_sdt, +drug_discfl,drug_discst, +drug_discrsshould be the same.
+In the targeted population, the
armvariable in +adsldataset should have noNAs or no strings +with white spaces only.
+Although not required, a good practice is to convert all +character variables to use into factors.
+-
+
adsl: ADSL dataset, dataframe.
+
+ -
+
adae: ADAE dataset, dataframe.
+
+ -
+
arm: Arm variable, character."TRT01A"by +default.
+
+ -
+
dose_adjust_flags: A vector of characters, names of +logic variables that indicate if AEs lead to any dose adjustment, such +as a drug discontinuation, dose interruption and reduction. The logic +variable only has TRUE/FALSE. Default is +NA.
+
+ -
+
dose_adjust_labels: A vector of characters, labels to +be displayed in the table for the logic variables in +dose_adjust_labels, e.g.: AE leading to drug +discontinuation from drug X. Default isNA.
+
+ -
+
gr34_highest_grade_only: Logical value. Default is +TRUE, such that only patients with the highest AE grade as 3 or 4 are +included in the counts of the"Grade 3-4 AE"and +"Treatment-related Grade 3-4 AE"; set it to +FALSEif you want to include patients with the highest AE +grade as 5 in the counts.
+ - adae dataset has following variables:
+
+ -
+
USUBJID: Character.
+
+ -
+
STUDYID: Character.
+ -
+
AEDECOD: Character.
+
+ -
+
AEBODSYS: Character.
+
+ -
+
ATOXGR: Numeric, 1, 2, 3, 4 and 5.
+
+ -
+
AESER: Character,'Y'and +'N'.
+
+ -
+
AEREL: Character,'Y'and +'N'.
+
+ -
+
SAFFL: Character,'Y'and +'N'.
+
+ -
+
ANL01FL: Character,'Y'and +'N'.
+
+ -
+
TRTEMFL: Character,'Y'and +'N'.
+ -
+
armvariable: Character.
+
+ -
+
dose_adjust_flagsvariable: logical,TRUE+andFALSE.
+ - The length of
dose_adjust_flagsis equal to the length +ofdose_adjust_labels.
+
+ - In the targeted population, the
armvariable in the +adsl and adae should 1) have noNAs or no strings with +whitespaces only; 2) have the same treatment arm levels.
+
+ - Although not required, a good practice is to convert all character +variables to use into factors. +
-
+
adsl: ADSL dataset, dataframe.
+
+ -
+
adae: ADAE dataset, dataframe.
+
+ -
+
arm: Arm variable, character. “TRT01A” by +default.
+
+ -
+
cutoff: Cutoff threshold, numeric.
+ -
+
adaedataset has following variables:
+ -
+
USUBJID: Character.
+ -
+
STUDYID: Character.
+ -
+
AEDECOD: Character.
+ -
+
AEBODSYS: Character.
+ -
+
ATOXGR: Numeric, 1, 2, 3, 4 and 5.
+ -
+
ANL01FL: Character,"Y"and +"N".
+ -
+
RACE: do we need to keep them in the +code +
+ -
+
COUNTRY: do we need to keep them in the +code +
+ -
+
armvariable: Character.
+ - In the targeted population, the
armvariable in +adslandadaeshould
+ - have no NAs or no strings with whitespaces only;
+
+ - have the same treatment arm levels.
+
+ - Although not required, a good practice is to convert all character +variables to use into factors. +
+
+
+
+
+
+
+
+
+
+
+
+
+
diff --git a/main/articles/tutorial.html b/main/articles/tutorial.html
new file mode 100644
index 00000000..9e98c128
--- /dev/null
+++ b/main/articles/tutorial.html
@@ -0,0 +1,847 @@
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ Articles
+
+
+ All vignettes
+ + +
+
+
+
+
+
+
+
+
+
+
+
+
+
+
diff --git a/main/authors.html b/main/authors.html
index 6031dd71..b0469111 100644
--- a/main/authors.html
+++ b/main/authors.html
@@ -17,7 +17,7 @@
@@ -25,6 +25,16 @@
+
+
+
+
+
+
+
+
+
+
+
+ SREP slide automation
+with
+ Joe Zhu
+(
+
+
+
+
+
+
+
+
+
+
+Tutorial (Beginner)
+SREP slide automation
+with AutoslideR
+
+ Joe Zhu
+(zhus31)
+
+
+
+ tutorial.Rmd
+1. Set up
+
+1. Set up autoslideR for SREP
+
+-
+
.
+├── atezo.pptx
+├── filters.yml
+├── metadata.yml
+├── programs
+│ ├── immotion010_data_process.R
+│ └── run_slides.R
+└── spec.yml-
+
+2. An
+
+
+2. An autoslideR step by step guide
+
+In the showcase below, we use the datasets from IMmotion010.
+
+rm(list = ls())
+library("rsvg")
+library("nestcolor")
+library("autoslideR")
+library("dplyr")
+library("rice")
+entiPath <- "root/clinical_studies/RO5541267/CDT30114/WO39210/data_analysis/CSRPrimary/prod/outdata_vad/"
+vad_list <- c("adsl", "adae")
+data_o <- rice_read(setNames(paste0(entiPath, vad_list, ".sas7bdat"), vad_list))
+rice_session_close()
+# add fake data for trt discontinuation (for showcase trt disposition table only)
+set.seed(123)
+data_o$adsl <- data_o$adsl %>%
+ mutate(
+ EOTSTT = sample(c("COMPLETED", "DISCONTINUED", "ONGOING"), 778, # The study has 778 patients in total
+ replace = TRUE, prob = c(0.5, 0.3, 0.2)
+ ),
+ DCSREAS = ifelse(EOTSTT == "DISCONTINUED", DCSREAS, NA)
+ )
+
+ADSL <- data_o$adsl
+ADAE <- data_o$adae
+2.1
+
+2.1 t_dm_slide demographic table
+
+
+
+2.1.1 Arguments +
+-
+
+
+2.1.2 Pre-conditions +
+-
+
-
+
-
+
+
+
+2.1.3 Example R code and output preview +
+
+# process data
+ADSL_m <- ADSL %>%
+ mutate_at(
+ c("TRT01A", "SEX"),
+ ~ sas_na(as.character(.))
+ ) %>%
+ mutate(
+ TRT01A = factor(TRT01A, levels = c("PLACEBO", "ATEZOLIZUMAB")),
+ SEX = factor(SEX, levels = c("M", "F"), labels = c("Male", "Female"))
+ ) %>%
+ var_relabel(
+ AAGE = "Age",
+ SEX = "Sex"
+ )
+
+t_dm_slide(ADSL_m,
+ vars = c("AAGE", "SEX"),
+ arm = "TRT01P",
+ stats = c("median", "range", "count_fraction")
+) %>% slides_preview()
+
+Demographic Table
+
+2.2
+
+2.2 t_ds_trt_slide treatment disposition table
+
+
+
+2.2.1 Arguments +
+-
+
+
+2.2.2 Pre-conditions +
+-
+
+
+
+2.2.3 Example R code and output preview +
+
+# process data
+ADSL_m <- ADSL %>%
+ mutate_at(
+ c("TRT01A", "EOTSTT", "DCSREAS"),
+ ~ sas_na(as.character(.))
+ ) %>%
+ mutate(
+ TRT01A = factor(TRT01A, levels = c("PLACEBO", "ATEZOLIZUMAB")),
+ EOSSTT = as.factor(toupper(EOTSTT)),
+ DTRTFL = as.factor(case_when(
+ EOTSTT == "DISCONTINUED" ~ "Y",
+ TRUE ~ "N"
+ ))
+ )
+
+t_ds_trt_slide(ADSL_m,
+ arm = "TRT01A",
+ drug_vars = c("ATEZOLIZUMAB", "PLACEBO"),
+ drug_names = c("Atezolizumab", "Placebo"),
+ drug_sdt = c("TRTSDT", "TRTSDT"),
+ drug_discfl = c("DTRTFL", "DTRTFL"),
+ drug_discst = c("EOTSTT", "EOTSTT"),
+ drug_discrs = c("DCSREAS", "DCSREAS"),
+) %>% slides_preview()
+
+Treatment Disposition Table
+
+2.3
+
+2.3 t_ae_summary_slide summary of safety
+
+
+
+2.3.1 Arguments +
+-
+
+
+2.3.2 Pre-conditions +
+-
+
-
+
-
+
+
+
+2.3.3 Example R code and output preview +
+
+# process data
+ADAE_m <- ADAE %>%
+ mutate(
+ TRT01A = sas_na(factor(TRT01A,
+ levels = c("PLACEBO", "ATEZOLIZUMAB")
+ )),
+ withdraw_flg = ifelse(AEACN == "DRUG WITHDRAWN", TRUE,
+ FALSE
+ ),
+ interrup_flg = ifelse(AEACN == "DRUG INTERRUPTED", TRUE,
+ FALSE
+ )
+ )
+
+t_ae_summ_slide(
+ adsl = ADSL_m, adae = ADAE_m,
+ arm = "TRT01A",
+ dose_adjust_flags = c("withdraw_flg", "interrup_flg"),
+ dose_adjust_labels = c("DRUG WITHDRAWN", "DRUG INTERRUPTED"),
+ gr34_highest_grade_only = TRUE
+) %>% slides_preview()
+
+Adverse event summary Table
+
+2.4
+
+2.4 t_ae_pt_slide summary of AEs
+
+
+
+2.4.1 Arguments +
+-
+
+
+2.4.2 Pre-conditions +
+-
+
-
+
-
+
-
+
-
+
+
+
+2.4.3 Example R code and output preview +
+
+ADAE_m <- ADAE %>%
+ mutate(TRT01A = sas_na(factor(TRT01A,
+ levels = c("PLACEBO", "ATEZOLIZUMAB")
+ )))
+t_ae_pt_slide(
+ adsl = ADSL_m,
+ adae = ADAE_m,
+ arm = "TRT01A"
+) %>% slides_preview()
+
+Adverse event Table
+
+
+
+ 3. Slide automation driven by human readable yaml files +
+The following code specify the output slide format and file paths +mentioned in the above section.
+
+# load filters
+filters::load_filters(file.path("IMmotion010", "filters.yml"), overwrite = TRUE)
+
+# create a table format with white background and grey header
+imc_format <- function(ft, ...) {
+ ft %>% autoslider_format(
+ odd_header = "grey",
+ odd_body = "white",
+ even_body = "white",
+ ...
+ )
+}
+
+# set up the spec file, slide theme template and output file
+spec_file <- file.path("IMmotion010", "spec.yml")
+deck_template <- file.path("IMmotion010", "atezo.pptx")
+outfile <- file.path("IMmotion010", "IMmotion010_demo.pptx")All examples below are referring to the above section, and we apply
+the “SE” filter to all following output
+3.1
+
+3.1 t_dm_slide demographic table
+
+
+
+
+3.1.1 Example Spec.yml +
+- program: t_dm_slide
+ titles: Demographics and Baseline Characteristics
+ footnotes: ''
+ paper: L6
+ suffix: SE
+ args:
+ arm: "TRT01P"
+ vars: ["AAGE", "SEX"]
+ stats: ["median", "range", "count_fraction"]User can specify the variables to be summarized using
+vars. In this example, we summarize age, sex. For
+continuous variables, summary statistics are median and range; for
+character/factor variables, summary statistic is count (percent).
+
+
+3.1.2 Example R code and output +
+
+# process data
+ADSL_m <- ADSL %>%
+ mutate_at(
+ c("TRT01A", "SEX"),
+ ~ sas_na(as.character(.))
+ ) %>%
+ mutate(
+ TRT01A = factor(TRT01A, levels = c("PLACEBO", "ATEZOLIZUMAB")),
+ SEX = factor(SEX, levels = c("M", "F"), labels = c("Male", "Female"))
+ ) %>%
+ var_relabel(
+ AAGE = "Age",
+ SEX = "Sex"
+ )
+
+# execute spec to generate slide
+study_spec <- spec_file %>% read_spec()
+outputs <- study_spec %>%
+ filter_spec(., program %in% c("t_dm_slide")) %>%
+ generate_outputs(datasets = list(adsl = ADSL_m)) %>%
+ decorate_outputs(version_label = NULL) %>%
+ generate_slides(outfile = outfile, template = deck_template, table_format = imc_format)
+
+Demographic Table
+
+3.2
+
+3.2 t_ds_trt_slide treatment disposition table
+
+
+
+
+3.2.1 Example Spec.yml +
+- program: t_ds_trt_slide
+ titles: Disposition
+ footnotes: ''
+ paper: L6
+ suffix: SE
+ args:
+ arm: "TRT01A"
+ colcount: FALSE
+ drug_vars: ["ATEZOLIZUMAB", "PLACEBO"]
+ drug_names: ["Atezolizumab", "Placebo"]
+ drug_sdt: ["TRTSDT", "TRTSDT"]
+ drug_discfl: ["DTRTFL", "DTRTFL"]
+ drug_discst: ["EOTSTT", "EOTSTT"]
+ drug_discrs: ["DCSREAS", "DCSREAS"]
+
+
+3.2.2 Example R code and output +
+
+# process data
+ADSL_m <- ADSL %>%
+ mutate_at(
+ c("TRT01A", "EOTSTT", "DCSREAS"),
+ ~ sas_na(as.character(.))
+ ) %>%
+ mutate(
+ TRT01A = factor(TRT01A, levels = c("PLACEBO", "ATEZOLIZUMAB")),
+ EOSSTT = as.factor(toupper(EOTSTT)),
+ DTRTFL = as.factor(case_when(
+ EOTSTT == "DISCONTINUED" ~ "Y",
+ TRUE ~ "N"
+ ))
+ )
+
+# execute spec to generate slide
+study_spec <- spec_file %>% read_spec()
+outputs <- study_spec %>%
+ filter_spec(., program %in% c("t_ds_trt_slide")) %>%
+ generate_outputs(datasets = list(adsl = ADSL_m)) %>%
+ decorate_outputs(version_label = NULL) %>%
+ generate_slides(outfile = outfile, template = deck_template, table_format = imc_format)
+3.3
+
+
+3.3 t_ae_summary_slide summary of safety
+
+
+
+
+
+3.3.2 Example R code and output +
+
+ADAE_m <- ADAE %>%
+ mutate(
+ TRT01A = sas_na(factor(TRT01A,
+ levels = c("PLACEBO", "ATEZOLIZUMAB")
+ )),
+ withdraw_flg = ifelse(AEACN == "DRUG WITHDRAWN", TRUE,
+ FALSE
+ ),
+ interrup_flg = ifelse(AEACN == "DRUG INTERRUPTED", TRUE,
+ FALSE
+ )
+ )
+# execute spec to generate slide
+study_spec <- spec_file %>% read_spec()
+outputs <- study_spec %>%
+ filter_spec(., program %in% c("t_ae_summ_slide")) %>%
+ generate_outputs(datasets = list(adsl = ADSL_m, adae = ADAE_m)) %>%
+ decorate_outputs(version_label = NULL) %>%
+ generate_slides(outfile = outfile, template = deck_template, table_format = imc_format)
+3.4
+
+3.4 t_ae_pt_slide summary of AEs
+
+
+
+
+
+3.4.1 Example Spec.yml +
+- program: t_ae_pt_slide
+ titles: Most common AEs (Experienced in at least 10% patients overall)
+ footnotes: ''
+ paper: L6
+ suffix: SE
+ args:
+ arm: "TRT01A"
+ cutoff: 10Note: This is an example spec for the table “most common AEs
+(Experienced in at least 10% patients overall)”. User can change the
+cutoff value as needed. t_ae_pt_slide serves as the
+backbone for many other AE tables. User can simply apply a different
+filter, or a combination of filters, to produce other AE tables. For
+example, user can use filter SE_REL to produce related AE
+tables, SE_SER for serious AE tables,
+SE_SER_REL for related serious AE tables. User can also
+define their own filters in filter.yml based on
+study-specific need. Below is another example spec for the table
+“Related G3-5 AEs (Experienced in at least 1% patients overall)”
- program: t_ae_pt_slide
+ titles: Related G3-5 AEs (Experienced in at least 1% patients overall)
+ footnotes: ''
+ paper: L6
+ suffix: SE_REL_G35
+ args:
+ arm: "TRT01A"
+ cutoff: 1where the filter entries were defined as
+ +
+
+
+3.4.2 Example R code and output +
+
+ADAE_m <- ADAE %>%
+ mutate(TRT01A = sas_na(factor(TRT01A,
+ levels = c("PLACEBO", "ATEZOLIZUMAB")
+ )))
+# execute spec to generate slide
+study_spec <- spec_file %>% read_spec()
+outputs <- study_spec %>%
+ filter_spec(., program %in% c("t_ae_pt_slide")) %>%
+ generate_outputs(datasets = list(adsl = ADSL_m, adae = ADAE_m)) %>%
+ decorate_outputs(version_label = NULL) %>%
+ generate_slides(outfile = outfile, template = deck_template, table_format = imc_format)
