merge-cli 3.5.1__tar.gz → 3.7__tar.gz

merge_cli-3.7/MANIFEST.in

@@ -0,0 +1 @@
+recursive-include merge_cli/data/models *.pkl *.py .gitkeep

merge_cli-3.7/PKG-INFO

@@ -0,0 +1,88 @@
+Metadata-Version: 2.4
+Name: merge-cli
+Version: 3.7
+Summary: MERGE variant pathogenicity prediction CLI with precomputed VCF cache and local/remote model integrations
+Author: MERGE Team
+License-Expression: LicenseRef-Proprietary
+Project-URL: Homepage, https://merge.fanglab.cn
+Keywords: bioinformatics,variant,pathogenicity,genomics,cli
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Environment :: Console
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+Requires-Dist: click>=8.1
+Requires-Dist: rich>=13.0
+Requires-Dist: requests>=2.31
+Requires-Dist: pysam>=0.22
+Requires-Dist: numpy>=1.26
+Requires-Dist: pandas>=2.1
+Requires-Dist: scikit-learn>=1.4
+Requires-Dist: joblib>=1.3
+Requires-Dist: matplotlib>=3.8
+Requires-Dist: shap>=0.44
+
+# merge-cli 3.7
+
+MERGE variant pathogenicity prediction CLI.
+
+## Highlights
+
+- Remote and local prediction modes.
+- Local precomputed VCF cache lookup for hg38/hg19 coding and splicing variants.
+- Local/remote model switches for AlphaGenome, HyenaDNA, NT, Evo2, Enformer, GENERATOR, GENERATOR-v2, and NT-v2.
+- Bundled MERGE ensemble model files.
+
+## Quick Start
+
+```bash
+pip install merge-cli
+merge --help
+merge predict --chrom chr1 --pos 69428 --ref T --alt G --genome hg38
+```
+
+## Precomputed VCF Cache
+
+```bash
+merge precomputed configure --data-dir /path/to/precomputed
+merge precomputed status
+merge precomputed download --genome all --variant-type all
+```
+
+Expected files include `coding_merged.vcf.gz`, `splicing_merged.vcf.gz`, their hg19 counterparts, and `.tbi` indices.
+
+## GENERATOR / GENERATOR-v2 / NT-v2
+
+These models use one external service by default at `http://127.0.0.1:18004`.
+
+```bash
+merge generator-ntv2 configure --url http://127.0.0.1:18004
+merge generator-ntv2 status
+merge predict --chrom chr1 --pos 69428 --ref T --alt G --generator --generator-v2 --nt-v2
+```
+
+Local mode uses the same service URL:
+
+```bash
+merge predict --local --chrom chr1 --pos 69428 --ref T --alt G --ensemble-type coding --generator --generator-v2 --nt-v2
+```
+
+Service contract:
+
+- `GET /health`
+- `POST /predict` with `chrom`, `pos`, `ref`, `alt`, `genome_version`, `models`
+- `POST /batch_predict` with `variants`, `models`
+
+Model keys: `generator`, `generator_v2`, `nt_v2`.
+
+## Enformer
+
+```bash
+merge enformer configure --url http://localhost:5004
+merge enformer status
+```
+
+If local services are unavailable, errors are reported under `prediction.errors` and the CLI continues returning available results.

merge_cli-3.7/README.md

@@ -0,0 +1,61 @@
+# merge-cli 3.7
+
+MERGE variant pathogenicity prediction CLI.
+
+## Highlights
+
+- Remote and local prediction modes.
+- Local precomputed VCF cache lookup for hg38/hg19 coding and splicing variants.
+- Local/remote model switches for AlphaGenome, HyenaDNA, NT, Evo2, Enformer, GENERATOR, GENERATOR-v2, and NT-v2.
+- Bundled MERGE ensemble model files.
+
+## Quick Start
+
+```bash
+pip install merge-cli
+merge --help
+merge predict --chrom chr1 --pos 69428 --ref T --alt G --genome hg38
+```
+
+## Precomputed VCF Cache
+
+```bash
+merge precomputed configure --data-dir /path/to/precomputed
+merge precomputed status
+merge precomputed download --genome all --variant-type all
+```
+
+Expected files include `coding_merged.vcf.gz`, `splicing_merged.vcf.gz`, their hg19 counterparts, and `.tbi` indices.
+
+## GENERATOR / GENERATOR-v2 / NT-v2
+
+These models use one external service by default at `http://127.0.0.1:18004`.
+
+```bash
+merge generator-ntv2 configure --url http://127.0.0.1:18004
+merge generator-ntv2 status
+merge predict --chrom chr1 --pos 69428 --ref T --alt G --generator --generator-v2 --nt-v2
+```
+
+Local mode uses the same service URL:
+
+```bash
+merge predict --local --chrom chr1 --pos 69428 --ref T --alt G --ensemble-type coding --generator --generator-v2 --nt-v2
+```
+
+Service contract:
+
+- `GET /health`
+- `POST /predict` with `chrom`, `pos`, `ref`, `alt`, `genome_version`, `models`
+- `POST /batch_predict` with `variants`, `models`
+
+Model keys: `generator`, `generator_v2`, `nt_v2`.
+
+## Enformer
+
+```bash
+merge enformer configure --url http://localhost:5004
+merge enformer status
+```
+
+If local services are unavailable, errors are reported under `prediction.errors` and the CLI continues returning available results.

merge_cli-3.7/merge_cli/__init__.py

@@ -0,0 +1 @@
+__version__ = "3.7"

{merge_cli-3.5.1 → merge_cli-3.7}/merge_cli/api.py

@@ -20,6 +20,14 @@ def _headers() -> dict:
     return {"Accept": "application/json"}
 
 
+def _session() -> requests.Session:
+    session = requests.Session()
+    # The MERGE endpoint is fixed and reachable directly. Some cluster nodes set
+    # HTTP(S)_PROXY values that break Python TLS while curl still works.
+    session.trust_env = False
+    return session
+
+
 # ─────────────────────────────────────────────────────────────
 # 1. 单变异预测  POST /predict/
 # ─────────────────────────────────────────────────────────────
@@ -34,7 +42,11 @@ def predict_single(
     use_gpn_msa:       bool = True,
     use_popeve:        bool = True,
     use_evo2:          bool = True,
-    use_evo1:          bool = True,
+    use_enformer:      bool = True,
+    use_generator:     bool = False,
+    use_generator_v2:  bool = False,
+    use_nt_v2:         bool = False,
+    use_evo1:          bool = True
 ) -> dict:
     payload = {
         "chr": chrom, "pos": pos, "ref": ref, "alt": alt,
@@ -47,9 +59,13 @@ def predict_single(
         "use_gpn_msa":       str(use_gpn_msa).lower(),
         "use_popeve":        str(use_popeve).lower(),
         "use_evo2":          str(use_evo2).lower(),
+        "use_enformer":      str(use_enformer).lower(),
+        "use_generator":     str(use_generator).lower(),
+        "use_generator_v2":  str(use_generator_v2).lower(),
+        "use_nt_v2":         str(use_nt_v2).lower(),
         "use_evo1":          str(use_evo1).lower(),
     }
-    resp = requests.post(
+    resp = _session().post(
         f"{_BASE}/predict/",
         json=payload,
         headers=_headers(),
@@ -89,7 +105,11 @@ def submit_batch(
     use_gpn_msa:       bool = True,
     use_popeve:        bool = True,
     use_evo2:          bool = True,
-    use_evo1:          bool = True,
+    use_enformer:      bool = True,
+    use_generator:     bool = False,
+    use_generator_v2:  bool = False,
+    use_nt_v2:         bool = False,
+    use_evo1:          bool = True
 ) -> dict:
     data = {
         "notify_email":    email,
@@ -102,10 +122,14 @@ def submit_batch(
         "use_gpn_msa":       str(use_gpn_msa).lower(),
         "use_popeve":        str(use_popeve).lower(),
         "use_evo2":          str(use_evo2).lower(),
+        "use_enformer":      str(use_enformer).lower(),
+        "use_generator":     str(use_generator).lower(),
+        "use_generator_v2":  str(use_generator_v2).lower(),
+        "use_nt_v2":         str(use_nt_v2).lower(),
         "use_evo1":          str(use_evo1).lower(),
     }
     with open(vcf_path, "rb") as f:
-        resp = requests.post(
+        resp = _session().post(
             f"{_BASE}/api/submit_batch_job/",
             data=data,
             files={"vcf_file": (vcf_path.split("/")[-1], f)},
@@ -120,7 +144,7 @@ def submit_batch(
 # 4. 查询批量任务状态  GET /api/batch_job/<job_id>/
 # ─────────────────────────────────────────────────────────────
 def get_batch_status(job_id: str) -> dict:
-    resp = requests.get(
+    resp = _session().get(
         f"{_BASE}/api/batch_job/{job_id}/",
         headers=_headers(),
         timeout=30,