NIST/Human/msp_to_mgf.py

# -*- coding: utf-8 -*-
import re
from pathlib import Path
from openpyxl import Workbook

# =========================================================
# 1) 路径配置：改这里即可
#    INPUT_DIR : 你截图里的 extracted 目录（存放 .msp）
#    OUTPUT_DIR: 输出 .mgf 的目录（会自动创建）
#    EXCEL_PATH: 输出统计表
# =========================================================
INPUT_DIR = Path("/Users/guanmumu/Desktop/Data/Human/extracted")
OUTPUT_DIR = Path("/Users/guanmumu/Desktop/Data/Human/mgf")
OUTPUT_DIR.mkdir(parents=True, exist_ok=True)

EXCEL_PATH = OUTPUT_DIR / "msp_to_mgf_summary.xlsx"

# 进度打印（大文件可调）
PROGRESS_EVERY_SPECTRA = 50000

# =========================================================
# 2) MSP 字段解析函数（保持和你单文件脚本一致）
# =========================================================
def parse_sequence_from_name(name_val: str) -> str:
    # Name: SEQUENCE/charge_...
    return name_val.split("/", 1)[0].strip()

def parse_charge_from_name(name_val: str):
    m = re.search(r"/(\d+)", name_val)
    return int(m.group(1)) if m else None

def parse_mz_exact_from_comment(line: str):
    # Comment: ... Mz_exact=536.5844 ...
    m = re.search(r"\bMz_exact=([0-9]*\.?[0-9]+)", line)
    return float(m.group(1)) if m else None

# =========================================================
# 3) 转换单个 MSP -> MGF，并返回统计信息
#    - Spectra count: Num peaks 次数
#    - Unique peptides: 对 Name: 的 seq 去重（与你单文件脚本一致）
# =========================================================
def convert_one_msp(msp_path: Path, mgf_path: Path):
    spectra_count = 0
    unique_peptides = set()

    with msp_path.open("r", encoding="utf-8", errors="replace") as fin, \
         mgf_path.open("w", encoding="utf-8") as fout:

        name_val = None
        seq = None
        charge = None
        mw = None               # 你要求用作 PEPMASS
        precursor_mz = None     # 仅用于辅助记录（Mz_exact）

        for line in fin:
            line = line.rstrip("\n")

            if line.startswith("Name:"):
                name_val = line.split("Name:", 1)[1].strip()
                seq = parse_sequence_from_name(name_val)
                charge = parse_charge_from_name(name_val)
                unique_peptides.add(seq)

            elif line.startswith("MW:"):
                try:
                    mw = float(line.split("MW:", 1)[1].strip())
                except Exception:
                    mw = None

            elif line.startswith("Comment:"):
                precursor_mz = parse_mz_exact_from_comment(line)

            elif line.startswith("Num peaks:"):
                num_peaks = int(line.split("Num peaks:", 1)[1].strip())
                spectra_count += 1

                # ===== 写一个 MGF block =====
                fout.write("BEGIN IONS\n")

                if seq is not None and charge is not None:
                    fout.write(f"TITLE={seq}/z{charge}\n")
                elif seq is not None:
                    fout.write(f"TITLE={seq}\n")
                else:
                    fout.write(f"TITLE=Spectrum_{spectra_count}\n")

                # 你要求：PEPMASS = MW（不做换算）
                if mw is not None:
                    fout.write(f"PEPMASS={mw:.6f}\n")

                if charge is not None:
                    fout.write(f"CHARGE={charge}+\n")

                # 自定义字段
                if seq is not None:
                    fout.write(f"SEQ={seq}\n")

                # 可选：记录 Mz_exact（不影响你要求）
                if precursor_mz is not None:
                    fout.write(f"PRECURSOR_MZ={precursor_mz:.6f}\n")

                # 峰：每行 "m/z intensity"
                for _ in range(num_peaks):
                    peak_line = next(fin).strip()
                    if not peak_line:
                        continue
                    parts = peak_line.split()
                    if len(parts) >= 2:
                        fout.write(f"{parts[0]} {parts[1]}\n")

                fout.write("END IONS\n\n")

                # reset 当前谱状态（与你单文件脚本一致）
                name_val = None
                seq = None
                charge = None
                mw = None
                precursor_mz = None

                if spectra_count % PROGRESS_EVERY_SPECTRA == 0:
                    print(f"  ... {msp_path.name}: spectra={spectra_count}")

    return spectra_count, len(unique_peptides)

# =========================================================
# 4) 写 Excel（3列：文件名 / Spectra count / Unique peptides）
# =========================================================
def write_excel(excel_path: Path, rows):
    wb = Workbook()
    ws = wb.active
    ws.title = "summary"

    for r in rows:
        ws.append(list(r))

    ws.column_dimensions["A"].width = 55
    ws.column_dimensions["B"].width = 18
    ws.column_dimensions["C"].width = 20

    wb.save(excel_path)

# =========================================================
# 5) 批量处理
# =========================================================
def main():
    if not INPUT_DIR.exists():
        raise FileNotFoundError(f"INPUT_DIR not found: {INPUT_DIR}")

    msp_files = sorted(INPUT_DIR.glob("*.msp"))
    if not msp_files:
        print(f"No .msp files found in: {INPUT_DIR}")
        return

    print(f"Input dir : {INPUT_DIR}")
    print(f"Output dir: {OUTPUT_DIR}")
    print(f"Found {len(msp_files)} MSP files.")

    rows = [("file_name", "spectra_count", "unique_peptides")]

    for idx, msp_path in enumerate(msp_files, 1):
        mgf_path = OUTPUT_DIR / (msp_path.stem + ".mgf")

        print(f"\n[{idx}/{len(msp_files)}] Converting: {msp_path.name}")
        spectra_count, unique_pep = convert_one_msp(msp_path, mgf_path)

        print(f"  -> MGF: {mgf_path.name}")
        print(f"  Spectra count   : {spectra_count}")
        print(f"  Unique peptides : {unique_pep}")

        rows.append((msp_path.name, spectra_count, unique_pep))

    write_excel(EXCEL_PATH, rows)

    print("\n================= DONE =================")
    print(f"Excel summary: {EXCEL_PATH}")

if __name__ == "__main__":
    main()