fix unique psms ids issue

hot-fix for non-unique input files
Rappsilber-Laboratory · Mar 2, 2021 · c4d073c · c4d073c
1 parent c75b841
commit c4d073c
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 2 deletions.
diff --git a/sample_data/DSS_xisearch_fdr_CSM50percent_transfer_scx17to23_hsax2to9_minimal.csv b/sample_data/DSS_xisearch_fdr_CSM50percent_transfer_scx17to23_hsax2to9_minimal.csv
@@ -1,4 +1,4 @@
-FDR,Fasta1,Fasta2,LinkPos1,LinkPos2,PSMID,Peptide1,Peptide2,Run,fdrGroup,isTT,score,rp,scx,hsax,isTD,isDD
+fdr,Fasta1,Fasta2,LinkPos1,LinkPos2,PSMID,Peptide1,Peptide2,Run,fdrGroup,isTT,score,rp,scx,hsax,isTD,isDD
 0.0,SUCC_ECOLI Succinate--CoA ligase [ADP-forming] subunit beta OS=Escherichia coli (strain K12) OX=83333 GN=sucC PE=1 SV=1,SUCD_ECOLI Succinate--CoA ligase [ADP-forming] subunit alpha OS=Escherichia coli (strain K12) OX=83333 GN=sucD PE=1 SV=2,11,12,333321,VAEETPHLIHKVALDPLTGPMPYQGR,MGHAGAIIAGGKGTADEK,recal_B181130_09_HF_FW_IN_130_ECLP_DSS01_SCX22_hSAX06_rep1.mgf,Between ,True,27.165,2905.68822,22,6,False,False
 0.0,SUCC_ECOLI Succinate--CoA ligase [ADP-forming] subunit beta OS=Escherichia coli (strain K12) OX=83333 GN=sucC PE=1 SV=1,SUCD_ECOLI Succinate--CoA ligase [ADP-forming] subunit alpha OS=Escherichia coli (strain K12) OX=83333 GN=sucD PE=1 SV=2,11,12,1414170,VAEETPHLIHKVALDPLTGPMPYQGR,MoxGHAGAIIAGGKGTADEK,recal_B181203_16_HF_FW_IN_130_ECLP_DSS01_SCX23_hSAX06_rep2.mgf,Between ,True,26.019,2833.43868,23,6,False,False
 0.0,RS6_ECOLI 30S ribosomal protein S6 OS=Escherichia coli (strain K12) OX=83333 GN=rpsF PE=1 SV=1,RS18_ECOLI 30S ribosomal protein S18 OS=Escherichia coli (strain K12) OX=83333 GN=rpsR PE=1 SV=2,11,6,1625137,HAVTEASPMVKAK,DIATLKNYITESGK,recal_B181123_21_HF_FW_IN_130_ECLP_DSS01_SCX21_hSAX05_rep2.mgf,Between ,True,24.917,3094.394160000001,21,5,False,False

diff --git a/xirt/const.py b/xirt/const.py
@@ -192,7 +192,7 @@
   # They must be carefully adapted for each prediction task.
   # recommended to use sigmoid for fractions (SCX/hSAX) if ordinal regression method should be used
   hsax-activation: sigmoid
-  # columne where the fraction RT is in the CSV input
+  # column where the fraction RT is in the CSV input ("xx_ordinal" xx_
   hsax-column: hsax_ordinal
   # the number of unique / distinct values (e.g. fractions)
   hsax-dimension: 10

diff --git a/xirt/predictor.py b/xirt/predictor.py
@@ -397,6 +397,10 @@ def preprocess(matches_df, sequence_type="crosslink", max_length=-1, cl_residue=
     # sort to keep only highest scoring peptide from duplicated entries
     matches_df = matches_df.sort_values(by="score", ascending=False)
 
+    if len(matches_df["PSMID"].unique()) != matches_df.shape[0]:
+        logger.warning("PSMID column was not unique! Redundant PSMIDs were removed")
+        matches_df = matches_df.drop_duplicates("PSMID")
+
     logger.info("Reordering peptide sequences. (mode: {})".format(sequence_type))
 
     # generate columns to handle based on input data type