spec2function 0.1.1__py3-none-any.whl

Spec2Function/workflow.py

@@ -0,0 +1,233 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Any, Dict, Iterable, Optional, Tuple, Union
+import json
+
+import pandas as pd
+
+from .single_analysis import SingleSpectrumAnalyzer
+from .set_analysis import MetaboliteSetAnalyzer
+
+SpectrumInput = Union[str, Dict[str, Any]]
+CSVInput = Union[str, Path, pd.DataFrame]
+
+
+def parse_json_spectrum(json_input: SpectrumInput) -> Dict[str, Any]:
+    if isinstance(json_input, (str, bytes, bytearray)):
+        data = json.loads(json_input)
+    elif isinstance(json_input, dict):
+        data = json_input
+    else:
+        raise TypeError("json_input must be a JSON string or a dict.")
+
+    peaks = data.get("peaks") or []
+    if not peaks:
+        raise ValueError("No peaks found in JSON input.")
+
+    mz_values = []
+    intensity_values = []
+    for peak in peaks:
+        if not isinstance(peak, (list, tuple)) or len(peak) < 2:
+            continue
+        mz_values.append(float(peak[0]))
+        intensity_values.append(float(peak[1]))
+
+    if not mz_values:
+        raise ValueError("No valid m/z-intensity pairs found in JSON input.")
+
+    precursor_mz = float(data.get("precursor_mz") or 0.0)
+    return {
+        "mz": mz_values,
+        "intensity": intensity_values,
+        "precursor_mz": precursor_mz,
+    }
+
+
+def load_set_dataframe(data: CSVInput) -> pd.DataFrame:
+    if isinstance(data, pd.DataFrame):
+        return data.copy()
+    return pd.read_csv(Path(data))
+
+
+def filter_set_dataframe(
+    df: pd.DataFrame,
+    min_abs_logfc: float = 0.1,
+    max_pvalue: float = 0.05,
+) -> Tuple[pd.DataFrame, Dict[str, Any]]:
+    filtered = df.copy()
+    columns = list(filtered.columns)
+    logfc_col = next((c for c in columns if "logfc" in c.lower()), None)
+    pval_col = next(
+        (c for c in columns if "pval" in c.lower() or "p.value" in c.lower()),
+        None,
+    )
+
+    if logfc_col and min_abs_logfc and min_abs_logfc > 0:
+        filtered = filtered[filtered[logfc_col].abs() >= min_abs_logfc]
+
+    if pval_col and max_pvalue is not None and max_pvalue < 1:
+        filtered = filtered[filtered[pval_col] <= max_pvalue]
+
+    info = {
+        "total": len(df),
+        "kept": len(filtered),
+        "logfc_col": logfc_col,
+        "pval_col": pval_col,
+        "min_abs_logfc": min_abs_logfc,
+        "max_pvalue": max_pvalue,
+    }
+    return filtered, info
+
+
+def _filter_papers(papers: Iterable[Dict[str, Any]], selected_pmids: Optional[Iterable[str]]) -> list:
+    if not selected_pmids:
+        return list(papers or [])
+    selected = {str(pmid) for pmid in selected_pmids}
+    return [p for p in (papers or []) if str(p.get("pmid")) in selected]
+
+
+def _ensure_analyzer(
+    analyzer: Optional[MetaboliteSetAnalyzer],
+    project_root: Optional[Path],
+    *,
+    device=None,
+    enable_gpt_pubmed: bool = True,
+) -> MetaboliteSetAnalyzer:
+    if analyzer is not None:
+        return analyzer
+    if project_root is None:
+        raise ValueError("project_root is required when analyzer is not provided.")
+    return MetaboliteSetAnalyzer.create_from_spec2function_root(
+        project_root,
+        device=device,
+        enable_gpt_pubmed=enable_gpt_pubmed,
+    )
+
+
+@dataclass
+class MS2BioTextWorkflow:
+    analyzer: MetaboliteSetAnalyzer
+
+    @classmethod
+    def from_spec2function_root(
+        cls,
+        project_root: Path,
+        *,
+        device=None,
+        enable_gpt_pubmed: bool = True,
+    ) -> "MS2BioTextWorkflow":
+        analyzer = MetaboliteSetAnalyzer.create_from_spec2function_root(
+            project_root,
+            device=device,
+            enable_gpt_pubmed=enable_gpt_pubmed,
+        )
+        return cls(analyzer=analyzer)
+
+    def run_single(
+        self,
+        json_input: SpectrumInput,
+        *,
+        top_k: int = 10,
+        user_focus: Optional[str] = None,
+        selected_pmids: Optional[Iterable[str]] = None,
+        include_annotation: bool = True,
+    ) -> Dict[str, Any]:
+        spectrum = parse_json_spectrum(json_input)
+        result = self.analyzer.single_inference(
+            spectrum["mz"],
+            spectrum["intensity"],
+            precursor_mz=spectrum["precursor_mz"],
+            top_k=top_k,
+        )
+
+        if include_annotation:
+            papers = _filter_papers(result.get("papers", []), selected_pmids)
+            annotation = self.analyzer.generate_annotation(
+                retrieved_fragments=result.get("retrieved_fragments", []),
+                papers=papers,
+                user_focus=user_focus,
+            )
+            result["annotation"] = annotation
+        return result
+
+    def run_set(
+        self,
+        data: CSVInput,
+        *,
+        background_info: Optional[str] = None,
+        min_abs_logfc: float = 0.1,
+        max_pvalue: float = 0.05,
+        min_features: int = 5,
+    ) -> Dict[str, Any]:
+        df = load_set_dataframe(data)
+        filtered, info = filter_set_dataframe(df, min_abs_logfc, max_pvalue)
+
+        if len(filtered) < min_features:
+            return {
+                "error": f"Too few features selected ({len(filtered)})",
+                "filter": info,
+            }
+
+        result = self.analyzer.run_semi_supervised_analysis(
+            filtered, background_info=background_info
+        )
+        result["filter"] = info
+        return result
+
+
+def run_single(
+    json_input: SpectrumInput,
+    *,
+    project_root: Optional[Path] = None,
+    analyzer: Optional[MetaboliteSetAnalyzer] = None,
+    device=None,
+    enable_gpt_pubmed: bool = True,
+    top_k: int = 10,
+    user_focus: Optional[str] = None,
+    selected_pmids: Optional[Iterable[str]] = None,
+    include_annotation: bool = True,
+) -> Dict[str, Any]:
+    analyzer = _ensure_analyzer(
+        analyzer,
+        project_root,
+        device=device,
+        enable_gpt_pubmed=enable_gpt_pubmed,
+    )
+    workflow = MS2BioTextWorkflow(analyzer=analyzer)
+    return workflow.run_single(
+        json_input,
+        top_k=top_k,
+        user_focus=user_focus,
+        selected_pmids=selected_pmids,
+        include_annotation=include_annotation,
+    )
+
+
+def run_set(
+    data: CSVInput,
+    *,
+    project_root: Optional[Path] = None,
+    analyzer: Optional[MetaboliteSetAnalyzer] = None,
+    device=None,
+    enable_gpt_pubmed: bool = True,
+    background_info: Optional[str] = None,
+    min_abs_logfc: float = 0.1,
+    max_pvalue: float = 0.05,
+    min_features: int = 5,
+) -> Dict[str, Any]:
+    analyzer = _ensure_analyzer(
+        analyzer,
+        project_root,
+        device=device,
+        enable_gpt_pubmed=enable_gpt_pubmed,
+    )
+    workflow = MS2BioTextWorkflow(analyzer=analyzer)
+    return workflow.run_set(
+        data,
+        background_info=background_info,
+        min_abs_logfc=min_abs_logfc,
+        max_pvalue=max_pvalue,
+        min_features=min_features,
+    )

spec2function-0.1.1.dist-info/METADATA

@@ -0,0 +1,91 @@
+Metadata-Version: 2.4
+Name: spec2function
+Version: 0.1.1
+Summary: Deep learning model for MS2 data annotation
+Home-page: https://huggingface.co/cgxjdzz/ms2function-assets
+Author: User
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: torch
+Requires-Dist: transformers
+Requires-Dist: numpy
+Requires-Dist: pandas
+Requires-Dist: scikit-learn
+Requires-Dist: tqdm
+Requires-Dist: wandb
+Requires-Dist: huggingface_hub
+Dynamic: author
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: home-page
+Dynamic: license-file
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
+
+# Spec2Function
+
+Spec2Function provides MS2 spectrum annotation and metabolite set analysis powered by
+MS2BioText models. Large model/data assets are hosted on Hugging Face Hub and are
+downloaded automatically on first use.
+
+## Install
+
+```bash
+pip install spec2function
+```
+
+Or for local development:
+
+```bash
+pip install -e .
+```
+
+## Assets (Hugging Face Hub)
+
+Default asset repo: `cgxjdzz/spec2function-assets`
+
+The package looks for the following files:
+
+- `models/best_model.pth`
+- `models/config.json`
+- `data/hmdb_subsections_WITH_NAME.jsonl`
+- `data/all_jsonl_embeddings.pt`
+
+Environment overrides:
+
+- `MS2FUNCTION_ASSET_DIR` (use local assets directory)
+- `MS2FUNCTION_ASSET_REPO` (override HF repo id)
+- `HUGGINGFACE_HUB_TOKEN` or `HF_TOKEN` (for private repos)
+
+## Quickstart (single spectrum)
+
+```python
+from pathlib import Path
+from Spec2Function import run_single
+
+json_input = {
+    "peaks": [[100.1, 200.0], [150.2, 300.0]],
+    "precursor_mz": 250.3,
+}
+
+result = run_single(json_input, project_root=Path(r"d:\NTU\Spec2Function"))
+print(result)
+```
+
+## Quickstart (metabolite set)
+
+```python
+from pathlib import Path
+from Spec2Function import run_set
+
+result = run_set(
+    r"d:\path\to\your.csv",
+    project_root=Path(r"d:\NTU\Spec2Function"),
+)
+print(result)
+```

spec2function-0.1.1.dist-info/RECORD

@@ -0,0 +1,23 @@
+Spec2Function/MS2BioTextDataset.py,sha256=HRQwAhX4DWbMKZ1ImRBnoVnGaHGmhKxBsqoU92FEZ44,133704
+Spec2Function/__init__.py,sha256=ygARHwFRVtdc1w5k-wBtRykYy1myXcdvwI3zTj-0YBc,443
+Spec2Function/assets.py,sha256=0z2Ltca7hIF-fPJqS8MbPRJymKm0Z64pKTVqVOslG3Q,2390
+Spec2Function/biotext_processor.py,sha256=aqy04iupcT-2Xxm_kwCMs9S9TWJSdq2fn1IFPY5rBlA,18225
+Spec2Function/config.py,sha256=Bk1WAM3o35gBSfPnf4e2HYkMfqML1DO9G2JYAGgSv8M,4299
+Spec2Function/data_augmentation.py,sha256=v2q8JbVNbkDPjPdQRlhFoac5O79_K-y8dDNSNp1D3sY,11227
+Spec2Function/gpt_inference.py,sha256=9x68FluHNFf0zyTTj3om0pKdZQYq-vy21g0ofqbsjns,33776
+Spec2Function/llm_client.py,sha256=VdGwepR1YzBDM-C2Z98-c4mi6H6gff2OCz_EYsIP0yI,3747
+Spec2Function/model_manager.py,sha256=8OYdegqOnQcmqQNpZqe9KqJFexm3OhdD-SxGcjUgOz4,51141
+Spec2Function/pubmed.py,sha256=WXeRyueZnsEk55cXOqummQ3IBi6cKo3-uOZnE327FdY,8718
+Spec2Function/read_raw_data.py,sha256=J1kP326xQ5xjk2FR3uNj7rkSCveYd-IBvfCP0NbhCcU,5504
+Spec2Function/utils.py,sha256=EszXC5EJTxSiFLLyTo2e6V251Xogbr0yA4s0uQxgUvU,6286
+Spec2Function/workflow.py,sha256=UU9dZ8zq70yBaHsE7BlHGotJajG9mbHCE9k5dY4hSaA,6836
+Spec2Function/model/MS2BioText.py,sha256=ZPF3Ay13wE-oEHeU0V__HIy1WDJVVoOlFdA7jG5s9OU,20361
+Spec2Function/model/MSBERT.py,sha256=mJ5xSAXl9G65ONcR75675o6HiMWSwmLrkEL0gTMmV54,9873
+Spec2Function/model/__init__.py,sha256=yhAmaYM2zaLZC13-vwGcHzQEnoMvdAb6JTJZhd_anCw,1238
+Spec2Function/model/config.py,sha256=5OYH1ah7_pTY1bmHVgDvg6u4jbdeft-OJXp_bdZtOoo,8867
+Spec2Function/model/utils.py,sha256=y0rGKbZ_CXEUIo26nM4LZe76aWwto9WxqenwdqxvEL0,5768
+spec2function-0.1.1.dist-info/licenses/LICENSE,sha256=McZ8PZH1HyFW5b29gkn4VsVQ7pTeUw45Ls9Rs-mMMhU,1081
+spec2function-0.1.1.dist-info/METADATA,sha256=7hY1As9b5nXtoMzPbzv-c0WNocrBhUhDEeyZK9pQs0A,2127
+spec2function-0.1.1.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+spec2function-0.1.1.dist-info/top_level.txt,sha256=bMENaDm5SnPDfBwmi0N5uC7b-O8UZKzpHP3OZXQq9sU,14
+spec2function-0.1.1.dist-info/RECORD,,

spec2function-0.1.1.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (82.0.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

spec2function-0.1.1.dist-info/licenses/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 MS2Function contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

spec2function-0.1.1.dist-info/top_level.txt

@@ -0,0 +1 @@
+Spec2Function