Merge pull request #104 from hgb-bin-proteomics/feat/top-down-convert

Feat/top down convert

Author: michabirklbauer

convert/topdown_msf_to_msamanda.py

@@ -0,0 +1,122 @@
+#!/usr/bin/env python3
+#
+# /// script
+# requires-python = ">=3.12"
+# dependencies = [
+#   "pandas",
+#   "psm-utils"
+# ]
+# ///
+
+import sqlite3
+import argparse
+import pandas as pd
+from psm_utils.io import read_file
+
+__version = "1.0.0"
+__data = "2026-03-22"
+
+
+def read_msf(msf_file: str) -> dict[str, pd.DataFrame]:
+    conn = sqlite3.connect(msf_file)
+    proteoforms = pd.read_sql_query(
+        "SELECT * FROM TargetProteoformSpectrumMatchs", conn
+    )
+    modifications = pd.read_sql_query("SELECT * FROM FoundModifications", conn)
+    conn.close()
+    return {"proteoforms": proteoforms, "modifications": modifications}
+
+
+def to_msamanda(msf_file: str, verify: bool = True) -> int:
+    msf = read_msf(msf_file)
+    ##
+    mods = dict()
+    for i, row in msf["modifications"].iterrows():
+        name = str(row["Name"]).strip()
+        abbreviation = str(row["Abbreviation"]).strip()
+        mass = float(row["DeltaMonoisotopicMass"])
+        if name not in mods:
+            mods[name] = mass
+        if abbreviation not in mods:
+            mods[abbreviation] = mass
+    ##
+    title = list()
+    sequence = list()
+    modifications = list()
+    protein_accessions = list()
+    score = list()
+    mz = list()
+    charge = list()
+    rt = list()
+    filename = list()
+    id = list()
+    for i, row in msf["proteoforms"].iterrows():
+        title.append(
+            f"controllerType=0 controllerNumber=1 scan={row['FragmentationScans']}"
+        )
+        sequence.append(row["ModifiedSequence"])
+        modifications_str = ""
+        if not pd.isna(row["Modifications"]):
+            for mod in str(row["Modifications"]).split(";"):
+                loc = mod.split("(")[0].strip()
+                name = ")".join("(".join(mod.split("(")[1:]).split(")")[:-1]).strip()
+                mass = mods[name]
+                modifications_str += f"{loc}({name}|{mass}|variable);"
+        modifications.append(modifications_str.rstrip(";"))
+        if pd.isna(row["ParentProteinAccessions"]):
+            protein_accessions.append("sp|UNKNOWN")
+        else:
+            protein_accessions.append(row["ParentProteinAccessions"])
+        score.append(row["CScore"])
+        mz.append(row["MassOverCharge"])
+        charge.append(row["Charge"])
+        rt.append(row["RetentionTime"])
+        filename.append(row["SpectrumFileName"])
+        id.append(i)
+    amanda = pd.DataFrame(
+        {
+            "Title": title,
+            "Sequence": sequence,
+            "Modifications": modifications,
+            "Protein Accessions": protein_accessions,
+            "Amanda Score": score,
+            "m/z": mz,
+            "Charge": charge,
+            "RT": rt,
+            "Filename": filename,
+            "Id": id,
+        }
+    )
+    amanda.to_csv(f"{msf_file}.csv", sep="\t", index=False)
+    if verify:
+        psms = read_file(f"{msf_file}.csv", filetype="msamanda")
+        print(f"Successfully read {len(psms)} PSMs from file!")
+    return 0
+
+
+def main(argv=None) -> int:
+    parser = argparse.ArgumentParser(
+        prog="topdown_msf_to_msamanda.py",
+        description="Converts a MSF file with top-down proteomics results to MS Amanda format.",
+        epilog="(c) Bioinformatics Research Group, FH OÖ Campus Hagenberg, 2026",
+    )
+    parser.add_argument(
+        dest="msf",
+        help="MSF file to convert to MS Amanda format.",
+        type=str,
+    )
+    parser.add_argument(
+        "-c",
+        "--check",
+        dest="verify",
+        action="store_true",
+        help="Check PSMs with psm_utils.",
+    )
+    parser.add_argument("--version", action="version", version=__version)
+    args = parser.parse_args(argv)
+
+    return to_msamanda(args.msf, args.verify)
+
+
+if __name__ == "__main__":
+    exit(main())