BAWTO
Simple Trial Outcome Queries
library(tidyverse)
sr.tbls <- rsr::get_answers(81395) |> rsr::list_answers(concordance=T,collapse=T)
ae.tbl <- rsr::get_answers(70431) |> pluck("Adverse Events") # Adverse Events table
names(sr.tbls)## [1] "basic" "biomarker"
## [3] "relative outcomes" "Single Outcomes Ratio"
## [5] "single_population_outcomes"Results
Unstudied genes
review.bm <- sr.tbls$biomarker |> pull(biomarker.name) |> unlist(recursive = T) |> unique()
review.bm## [1] "BRCA1" "BRCA2" "ATM" "FANCA" "PALB2"
## [6] "BRIP1" "HDAC2" "ATR" "CHEK2" "MLH1"
## [11] "MRE11A" "NBN" "RAD51C" "BARD1" "CDK12"
## [16] "CHEK1" "FANCL" "PPP2R2A" "RAD51B" "RAD51D"
## [21] "RAD54L" "ETS fusion" "AR" "TP53" "PTEN"
## [26] "FOXA1" "ZBTB16" "NCOR1" "NCOR2" "PIK3CA"
## [31] "PIK3CB" "PIK3R1" "AKT1" "APC" "CTNNB1"
## [36] "RNF43" "RSPO2" "RB1" "CHD1" "SPOP"
## [41] "ZFHX3" "RAD51" "MLH3" "ERCC3" "MRE11"
## [46] "MDM2" "CDKN2A" "DKN1B" "MYC" "CCND1"
## [51] "MSH2" "MSH6" "FOXP1" "KMT2C" "KMT2D"
## [56] "KDM6A" "NKX3-1" "ARID1A" "ATRX" "FANCF"
## [61] "FANCG" "FANCM" "RAD50" "WRN"library(msigdbr)
gs <- msigdbr(species="Homo sapiens") |> filter(gs_name %in% c(
"KEGG_HOMOLOGOUS_RECOMBINATION",
"WP_HOMOLOGOUS_RECOMBINATION",
"REACTOME_HDR_THROUGH_HOMOLOGOUS_RECOMBINATION_HRR"
))
# How many HRR genes?
HRRgenes <- length(unique(gs$ensembl_gene))
# How many HRR genes unmeasured in our trials?
newgenes <- gs |> filter(!(human_gene_symbol %in% review.bm)) |>
group_by(entrez_gene,human_gene_symbol) |> count()Queryable data
Get the data
trial <- mutate(sr.tbls$basic,trial=unlist(short_name,recursive=T)) |> select(aid,trial)
spo <- sr.tbls$single_population_outcomes |> inner_join(trial,by="aid") |> modify(unlist) # single population outcomes
DT::datatable(spo)Longest reported median overall survival time
What was the longest reported median overall survival time in each trial?
spo |> filter(outcome_type=="OS") |>
mutate(outcome_value = as.numeric(outcome_value)) |>
group_by(trial) |> top_n(1,outcome_value) |> ungroup() |>
select(outcome_value,outcome_range,int_biomarker,bm_positive,int_bm_mod,trial) |>
arrange(desc(outcome_value))## Warning in mask$eval_all_mutate(quo): NAs introduced by coercion## # A tibble: 7 × 6
## outcome_value outcome_range int_biomarker bm_positive int_bm_mod trial
## <dbl> <chr> <chr> <chr> <chr> <chr>
## 1 32.3 28.4-NR none "NA" none Chin…
## 2 24.8 17.4-NC BRCA2 "" qualifyin… PROf…
## 3 22.7 17.4-29.4 none specified "NA" NA Saad…
## 4 19.1 NR BRCA1,BRCA2 "positive" qualifyin… PROf…
## 5 18.4 NA BRCA1,BRCA2 "qualifyin… altered Frie…
## 6 17.7 9.9-22.2 BRCA1,BRCA2 "positive" altered TOPA…
## 7 13.8 NR BRCA2,ATM,FANCA,CHEK… "positive" altered TOPA…Which PARP-i had the lowest and highest rate of grade 3+ anemia?
The adverse event counts table is loaded from cache below and is derived from ae.tbl which is imported from sysrev.com/p/70431 above. To see how the table is derived see github.com/insilica/bawto
AE.countDF <- readRDS(url("http://insilica.co.s3.amazonaws.com/bawto/resources/adverse_events.rds"))
AE.countDF |> filter(event=="anemia") |> filter(treatment_class=="PARPi",grade=="grade 3+") |> arrange(rate)## # A tibble: 10 × 16
## rg_name order treatment_class grade event patients affected rate tot
## <chr> <dbl> <chr> <fct> <chr> <dbl> <dbl> <dbl> <dbl>
## 1 chinnaiyan 2 PARPi grad… anem… 79 2 0.0253 5.51
## 2 TRITON2-1 12 PARPi grad… anem… 78 12 0.154 4.04
## 3 TOPARP-B 8 PARPi grad… anem… 49 10 0.204 0.469
## 4 saad 1 PARPi grad… anem… 71 15 0.211 5.21
## 5 PROfound-1 6 PARPi grad… anem… 256 55 0.215 4.45
## 6 PROfound-2 7 PARPi grad… anem… 256 58 0.227 4.66
## 7 TRITON2-2 13 PARPi grad… anem… 115 29 0.252 4.51
## 8 TOPARP-A-3… 9 PARPi grad… anem… 49 15 0.306 6.51
## 9 TOPARP-A-4… 10 PARPi grad… anem… 49 18 0.367 7.88
## 10 moran 0 PARPi grad… anem… NA NA NA 0.679
## # … with 7 more variables: tot_rate <dbl>, event_tot <int>, med_rate <dbl>,
## # mean_rate <dbl>, has_treatment <lgl>, ev.studies <int>, rate_bucket <fct>