Create 00_createSarcPDOSampleFile.py

add script for sarcpdo that downloads and combines genetic and transcriptomic sample IDs, merges and formats based on linkml schema

Author: RubyFore

build/sarcpdo/00_createSarcPDOSampleFile.py

@@ -0,0 +1,121 @@
+import synapseclient
+import pandas as pd
+import numpy as np
+import argparse
+import os
+
+def download_and_format_genetic_samples(synLoginObject):
+    """
+    Download and format samples that have genetic data in the Sarcoma PDO project via Synapse. 
+    
+    Parameters 
+    ----------
+    synLoginObject : synapseclient.Synapse
+        an object generated by a call to syn.login(PAT)
+
+    Returns
+    -------
+    pd.DataFrame
+        a dataframe containing formatted sample info 
+    """
+    # download genetic sample sheet, 15 rows
+    geneticSampleSheet = synLoginObject.tableQuery("select * from syn61894699")
+    geneticSampleDF = geneticSampleSheet.asDataFrame()
+    # subset and rename
+    genetic_samples = geneticSampleDF[['Sample_ID', 'Diagnosis']].rename({'Sample_ID':'common_name', 'Diagnosis':'cancer_type'}, axis=1)
+    genetic_samples['species'] = 'Homo sapiens(Human)'
+    genetic_samples['other_id_source'] = 'Synapse'
+    # append "_Tumor" to be more similar to RNAseq other_id's 
+    genetic_samples['other_id'] =  genetic_samples['common_name'].astype(str) + '_Tumor'
+    genetic_samples['other_id'] = genetic_samples['other_id'].str.replace("_2", "-2")
+    # assign all model types to tumor - is this correct?
+    genetic_samples['model_type'] = 'tumor'
+    # make empty column
+    genetic_samples['other_names'] = ''
+    # re-order columns
+    genetic_samples = genetic_samples[['other_id', 'common_name', 'other_id_source', 'other_names', 'cancer_type', 'species', 'model_type']]
+
+    return genetic_samples
+    
+
+def download_and_format_rna_samples(synLoginObject):
+    """
+    Download and format samples that have RNAseq data in the Sarcoma PDO project via Synapse. 
+
+    Parameters 
+    ----------
+    synLoginObject : synapseclient.Synapse
+        an object generated by a call to syn.login(PAT)
+
+    Returns
+    -------
+    pd.DataFrame
+        a dataframe containing formatted sample info 
+    """
+    # download rna sample sheet, 64 rows, 32 unique samples
+    rnaSampleSheet = synLoginObject.tableQuery("select * from syn61894657")
+    rnaSampleDF = rnaSampleSheet.asDataFrame()
+    # select and rename columns 
+    rna_samples = rnaSampleDF[['Sample_Name', 'individualID', 'diagnosis', 'experimentalCondition']].rename({"Sample_Name" : 'other_id', 'individualID':'common_name', 'diagnosis':'cancer_type', 'experimentalCondition':'model_type'}, axis=1)
+    # subset to only include one of the paired-end IDs
+    rna_samples = rna_samples[rna_samples['other_id'].str.contains("R1")]
+    # trimming 'Sample_Name' to only include sample ID and model_type 
+    rna_samples['other_id'] = rna_samples['other_id'].str.slice(0, -16)
+    # add 3 columns below
+    rna_samples['species'] = 'Homo sapiens(Human)'
+    rna_samples['other_id_source'] = 'Synapse'
+    rna_samples['other_names'] = ''
+    # re-order columns
+    rna_samples = rna_samples[['other_id', 'common_name', 'other_id_source', 'other_names', 'cancer_type', 'species', 'model_type']]
+    # replace abbreviation (MPNST) with full name
+    rna_samples.loc[rna_samples['cancer_type'] == 'MPNST', "cancer_type"] = 'Malignant peripheral nerve sheath tumor'
+    # sarc00095_tumor duplicated, Ewing Sarcoma (in RNA) vs. CIC-rearranged sarcoma in DNA
+    rna_samples.loc[rna_samples['cancer_type'] == 'Ewing sarcoma', "cancer_type"] = 'CIC-rearranged sarcoma'
+
+    # Sarc0101, 0137 - change name in RNA to full 'Dedifferentiated liposarcoma
+    rna_samples.loc[rna_samples['cancer_type'] == 'Dediff Liposarcoma', "cancer_type"] = 'Dedifferentiated liposarcoma'
+    # Sarc 0120 - change name in RNA to full 'Well-differentiated liposarcoma'    
+    rna_samples.loc[rna_samples['cancer_type'] == 'Well-diff Liposarcoma', "cancer_type"] = 'Well-differentiated liposarcoma'
+    # clean up 'model_type' by removing 'Thawed' prefix in some rows
+    modeltypeDF = rna_samples['model_type'].str.rsplit("_", expand =True)
+    modeltypeDF.loc[modeltypeDF[0] =="Thawed", [0]] = modeltypeDF[1] 
+    modeltypeDF[0] = modeltypeDF[0].str.lower()
+    rna_samples['model_type'] = modeltypeDF[0]
+    
+    return rna_samples
+
+    #def generate_samples_file(prev_samples_path):
+
+    # if prev_samples_path == "":
+        #maxval = 0
+   # else:
+    #    maxval = max(pd.read_csv(prev_samples_path).improve_sample_id)
+
+if __name__ == "__main__":
+    print('in main')
+    parser = argparse.ArgumentParser(description="This script handles downloading, processing and formatting of sample files for the Sarcoma PDO project into a single samplesheet")
+    print('in line 97')
+    parser.add_argument('-t', '--token', type=str, help='Synapse Token')
+
+    parser.add_argument("-p", '--prevSamples', nargs="?", type=str, default ="", const  = "", help = "Use this to provide previous sample file, will run sample file generation")
+
+    args = parser.parse_args()
+    print(args)
+    print("Logging into Synapse")
+    PAT = args.token
+    synObject = synapseclient.login(authToken=PAT)
+
+    rnaTable = download_and_format_rna_samples(synObject)
+    geneticTable = download_and_format_genetic_samples(synObject)
+    print()
+    merged = rnaTable.merge(geneticTable, how='outer')
+    
+    prev_max_improve_id = max(pd.read_csv(args.prevSamples).improve_sample_id)
+    merged['improve_sample_id'] = range(prev_max_improve_id+1, prev_max_improve_id+merged.shape[0]+1) 
+
+    merged.to_csv('~/Downloads/sarcpdo_samples.csv', index=False)
+
+        # validate with: linkml validate -s coderdata/schema/coderdata.yaml ~/Downloads/sarcpdo_samples.csv
+
+
+