addresses the issues in #251

mPnst and mpnstpdx code now build.

Author: sgosline

build/docker/Dockerfile.mpnst

@@ -1,4 +1,4 @@
-FROM r-base:4.3.2
+FROM r-base:4.3.3
 
 # Set environment to noninteractive
 ENV DEBIAN_FRONTEND=noninteractive

build/mpnst/00_sample_gen.R

@@ -7,14 +7,25 @@ library(dplyr)
 
 ##adding a command line argument
 args = commandArgs(trailingOnly=TRUE)
-if(length(args)!=2){
-    stop("Need a sample file and synapse token as argument. Rscript 00_sample_gen.R [samplefile] [synapse token]")
+if(length(args) > 1 ){
+    stop("Up to one argument is allowed. This is the filepath to the previously run samples file.")
+}
+
 
+if (length(args) == 0 || is.na(args[1]) || args[1] == "" || !file.exists(args[1])) {
+    orig_samples <- ""
+} else {
+    orig_samples <- fread(args[1])
 }
 
-orig_samples<-fread(args[1])
 
-synapser::synLogin(authToken=args[2])
+# Check if Synapse token is available from the environment
+synapse_token <- Sys.getenv("SYNAPSE_AUTH_TOKEN")
+if (synapse_token == "") {
+    stop("Error: SYNAPSE_AUTH_TOKEN environment variable is not set.")
+}
+
+synapser::synLogin(authToken=synapse_token)
 manifest<-synapser::synTableQuery("select * from syn53503360")$asDataFrame()|>
                                                              as.data.frame()
 
@@ -32,23 +43,18 @@ manifest<-synapser::synTableQuery("select * from syn53503360")$asDataFrame()|>
 ##first create samples for the original tumors
 tumorTable<-manifest|>
     dplyr::select(common_name='Sample')|>
-    dplyr::mutate(other_id_source='NF Data Portal',other_names='',cancer_type="Malignant peripheral nerve sheath tumor",species='Human',model_type='tumor')|>
+    dplyr::mutate(other_id_source='NF Data Portal',other_names='',cancer_type="Malignant peripheral nerve sheath tumor",species='Homo sapiens (Human)',model_type='tumor')|>
     tidyr::unite(col='other_id',c('common_name','model_type'),sep=' ',remove=FALSE)
 
 ##then create samples for the PDX
 sampTable<-manifest|>
     dplyr::select(common_name='Sample',MicroTissueDrugFolder)|>
-    dplyr::mutate(other_id_source='NF Data Portal',other_names='',cancer_type="Malignant peripheral nerve sheath tumor",species='Human',model_type='patient derived xenograft')|>
+    dplyr::mutate(other_id_source='NF Data Portal',other_names='',cancer_type="Malignant peripheral nerve sheath tumor",species='Homo sapiens (Human)',model_type='patient derived xenograft')|>
     tidyr::unite(col='other_id',c('common_name','model_type'),sep=' ',remove=FALSE)
 
 
-pdxmt<-manifest|>
-    dplyr::select(common_name='Sample',MicroTissueDrugFolder)|>
-    dplyr::mutate(other_id_source='NF Data Portal',other_names='',cancer_type="Malignant peripheral nerve sheath tumor",species='Human',model_type='organoid')|>
-    tidyr::unite(col='other_id',c('common_name','model_type'),sep=' ',remove=FALSE)
-
 ##third, generate a sample for the MTs if they were generated
-#pdxmt<-subset(sampTable,!is.na(MicroTissueDrugFolder))
+pdxmt<-subset(sampTable,!is.na(MicroTissueDrugFolder))
 pdxmt$model_type=rep('organoid',nrow(pdxmt))
 print(pdxmt)
 
@@ -58,7 +64,15 @@ main<-rbind(sampTable,pdxmt)|>
 
 #main <- fread("mpnst/NF_MPNST_samples.csv")
 #previous_aml <- fread(args[1])#"beatAML/beataml_samples.csv")
-max_id <- max(orig_samples$improve_sample_id)
+
+# If there is no previous samples file - start at 1, else, continue where the previous one left off.
+if (identical(orig_samples, "")) {
+    max_id <- 1  
+} else {
+    max_id <- max(orig_samples$improve_sample_id, na.rm = TRUE)
+}
+
+
 main$improve_sample_id <- seq(from = max_id + 1, length.out = nrow(main))
 
 #synapse_main <- fread("mpnst/synapse_NF-MPNST_samples.csv")

build/mpnst/03_get_drug_response_data.R

@@ -32,7 +32,7 @@ org_samps<-subset(samples_df,model_type=='organoid')
 
 ##now get the manifest from synapse
 manifest<-synapser::synTableQuery("select * from syn53503360")$asDataFrame()|>
-                                                             as.data.frame()|>
+                                                             as.data.table()|>
                                                              dplyr::rename(common_name='Sample')
 
 

build/mpnst/README.md

@@ -12,7 +12,7 @@ directory. Currently using the test files as input.
    `mpnst_samples.csv` file. This pulls from the latest synapse
    project metadata table.
    ```
-   docker run -v $PWD:/tmp -e SYNAPSE_AUTH_TOKEN=$SYNAPSE_AUTH_TOKEN mpnst sh build_samples.sh /tmp/build/build_test/test_samples.csv
+   docker run -v $PWD:/tmp -e -e SYNAPSE_AUTH_TOKEN=$SYNAPSE_AUTH_TOKEN mpnst sh build_samples.sh /tmp/build/build_test/test_samples.csv
    ```
 
 3. Pull the data and map it to the samples. This uses the metadata

build/mpnst/build_drugs.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
-set -euo pipefail
+#set -euo pipefail
 
-trap 'echo "Error on or near line $LINENO while executing: $BASH_COMMAND"; exit 1' ERR
+#trap 'echo "Error on or near line $LINENO while executing: $BASH_COMMAND"; exit 1' ERR
 
 echo "Running 02_get_drug_data.R with /tmp/mpnst_drugs.tsv and $1."
 Rscript 02_get_drug_data.R /tmp/mpnst_drugs.tsv $1

build/mpnst/requirements.r

@@ -4,8 +4,8 @@ install.packages('remotes')
 remotes::install_version('rjson', version = '0.2.21', repos = 'https://cloud.r-project.org')
 install.packages('synapser', repos = c('http://ran.synapse.org', 'https://cloud.r-project.org'))
 install.packages("dplyr")
-install.packages("data.table")
 install.packages("synapser", repos = c("http://ran.synapse.org", "https://cloud.r-project.org"))
+install.packages("data.table")
 install.packages("R.utils")
 install.packages("stringr")
-install.packages("tidyr")
+install.packages("tidyr")

build/utils/build_drug_desc.py

@@ -82,8 +82,9 @@ def main():
     cansmiles = [a for a in set(tab.canSMILES) if str(a)!='nan']
     #    isosmiles = list(set(tab.isoSMILES))
     morgs = smiles_to_fingerprint(cansmiles)
-
+#    print(morgs)
     ids = pd.DataFrame(tab[['improve_drug_id','canSMILES']]).drop_duplicates()
+#    print(ids)
     id_morg = ids.rename({"canSMILES":'smile'},axis=1).merge(morgs)[['improve_drug_id','structural_descriptor','descriptor_value']]
 
     mords = smiles_to_mordred(cansmiles,nproc=ncors)