Initial commit: digital-patients pipeline (clean, no large files)

Large reference/model files excluded from repo - to be staged to S3 or baked into Docker images.

app_filter/README.md

@@ -0,0 +1,5 @@
+Run command
+
+```
+vep --config vep.ini -i ${input_vcf} -o ${input_vcf.simpleName}_cosmic.vcf
+```

app_filter/filter_transcriptome.py

@@ -0,0 +1,56 @@
+from argparse import ArgumentParser
+import pandas as pd
+import numpy as np
+from pathlib import Path
+
+def make_argparser():
+    parser = ArgumentParser(
+        description="Filter Transcriptome using mutated and non-mutated TPM values, and MANE dataset"
+    )
+    parser.add_argument(
+        "--mutated",
+        type=Path,
+        required=True,
+        help="The .csv file with mutated TPM."
+    )
+    parser.add_argument(
+        "--non-mutated",
+        type=Path,
+        required=True,
+        help="The .csv file with base (non-mutated) TPM"
+    )
+    parser.add_argument(
+        "--mane-data",
+        type=Path,
+        required=True,
+        help="The .tsv file with MANE data"
+    )
+    return parser
+
+def main(mutated : Path, non_mutated : Path, mane_data : Path):
+    
+    name = mutated.stem
+    
+    df_mut = pd.read_csv(mutated)
+    df_nomut = pd.read_csv(non_mutated)
+    
+    df_mane = pd.read_csv(mane_data, sep="\t")
+    df_mane["ENSG"] = [i.split('.')[0] for i in df_mane["Ensembl_Gene"]]
+    df_mane_map = df_mane[['symbol','ENSG']]
+    
+    df_nomut_mapped = df_mane_map.merge(df_nomut, left_on='symbol', right_on='Unnamed: 0').dropna().drop_duplicates(subset=['symbol']).drop(["symbol", "Unnamed: 0"],axis=1)
+    
+    df_merge = df_nomut_mapped.merge(df_mut, on="ENSG", suffixes=("_nomut","_mut"))
+
+    cols = df_mut.columns[1:]
+    
+    df_res = pd.DataFrame({"ENSG": df_merge["ENSG"]} | {col:np.log2(df_merge[col + "_mut"] / df_merge[col + "_nomut"]) for col in cols})
+    
+    df_res.to_csv(f"{name}_log2fc.csv", index=None)
+
+
+if __name__ == "__main__":
+    parser = make_argparser()
+    args = parser.parse_args()
+    
+    main(mutated=args.mutated, non_mutated=args.non_mutated, mane_data=args.mane_data)

app_filter/vep.ini

@@ -0,0 +1,32 @@
+cache 1
+dir_cache  /data
+assembly 'GRCh38'
+offline 1
+force_overwrite 1
+custom file=/opt/vep/.vep/Cosmic_GenomeScreensMutant_v103_GRCh38.vcf.gz,short_name=COSMIC,format=vcf,type=exact,fields=LEGACY_ID%CNT
+
+# Functional impact predictions
+sift b # CALCULATE 1 - score
+polyphen b # THESE TWO CAN BE AVERAGED
+
+plugin REVEL,/opt/vep/.vep/REVEL/new_tabbed_revel_grch38.tsv.gz
+# TODO: CALCUALTE THEIR L2 NORM
+# plugin CADD,/opt/vep/.vep/CADD/GRCh38/whole_genome_SNVs.tsv.gz,/opt/vep/.vep/CADD/GRCh38/InDels.tsv.gz
+
+
+# Clinical significance
+check_existing
+clinvar
+
+# Allele frequencies
+af
+af_1kg
+af_gnomade
+af_gnomadg
+max_af
+
+# Additional useful flags
+numbers
+hgvs
+symbol
+canonical