geney 1.4.41__py3-none-any.whl → 1.4.44__py3-none-any.whl

geney/__init__.py

@@ -15,6 +15,7 @@ from .pipelines import (
     max_splicing_delta,
     oncosplice_pipeline_single_transcript,  # backwards compat
 )
+# from .samples import *
 
 __all__ = [
     "Mutation",
@@ -35,4 +36,5 @@ __all__ = [
 
 
 mut_id = 'KRAS:12:25227343:G:T'
-epistasis_id = 'KRAS:12:25227343:G:T|KRAS:12:25227344:A:T'
+epistasis_id = 'KRAS:12:25227343:G:T|KRAS:12:25227344:A:T'
+

geney/engines.py

@@ -6,9 +6,7 @@ import numpy as np
 
 # Lazy-loaded model containers (loaded automatically on first use)
 _pang_models = None
-_sai_models = None
 _pang_device = None
-_sai_device = None
 
 
 def _get_torch_device():
@@ -27,21 +25,6 @@ def _get_torch_device():
     return torch.device("cpu")
 
 
-def _get_tensorflow_device():
-    """Get the best available TensorFlow device."""
-    import sys
-    import tensorflow as tf
-
-    try:
-        if tf.config.list_physical_devices('GPU'):
-            return '/GPU:0'
-        elif sys.platform == 'darwin' and tf.config.list_physical_devices('MPS'):
-            return '/device:GPU:0'
-    except Exception:
-        pass
-    return '/CPU:0'
-
-
 def _load_pangolin_models():
     """Lazy load Pangolin models."""
     global _pang_models, _pang_device
@@ -75,49 +58,23 @@ def _load_pangolin_models():
     return _pang_models
 
 
-def _load_spliceai_models():
-    """Lazy load SpliceAI models."""
-    global _sai_models, _sai_device
+_OPENSPLICEAI_MODEL_DIR = None
 
-    if _sai_models is not None:
-        return _sai_models
+def _get_openspliceai_model_dir() -> str:
+    """Return the path to the OpenSpliceAI MANE 10000nt model directory."""
+    global _OPENSPLICEAI_MODEL_DIR
+    if _OPENSPLICEAI_MODEL_DIR is not None:
+        return _OPENSPLICEAI_MODEL_DIR
 
     import os
-    os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
-
-    import sys
-    import tensorflow as tf
-    from keras.models import load_model
-    from importlib import resources
 
-    import absl.logging
-    absl.logging.set_verbosity(absl.logging.ERROR)
+    # Models ship inside the package at geney/models/openspliceai-mane/10000nt
+    pkg_dir = os.path.dirname(os.path.abspath(__file__))
+    default = os.path.join(pkg_dir, 'models', 'openspliceai-mane', '10000nt')
+    default = os.path.normpath(default)
 
-    _sai_device = _get_tensorflow_device()
-    print(f"SpliceAI loading to {_sai_device}...")
-
-    _sai_models = []
-
-    try:
-        if sys.platform == 'darwin':
-            model_filenames = [f"models/spliceai{i}.h5" for i in range(1, 6)]
-            model_paths = [resources.files('spliceai').joinpath(f) for f in model_filenames]
-        else:
-            model_paths = [f"/tamir2/nicolaslynn/tools/SpliceAI/spliceai/models/spliceai{i}.h5"
-                           for i in range(1, 6)]
-
-        with tf.device(_sai_device):
-            for i, model_path in enumerate(model_paths):
-                try:
-                    model = load_model(str(model_path))
-                    _sai_models.append(model)
-                except Exception as e:
-                    print(f"Warning: Failed to load SpliceAI model {i+1}: {e}")
-    except Exception as e:
-        print(f"Error loading SpliceAI models: {e}")
-
-    print(f"SpliceAI loaded ({len(_sai_models)} models).")
-    return _sai_models
+    _OPENSPLICEAI_MODEL_DIR = os.environ.get('OPENSPLICEAI_MODEL_DIR', default)
+    return _OPENSPLICEAI_MODEL_DIR
 
 
 def pang_one_hot_encode(seq: str) -> np.ndarray:
@@ -142,25 +99,6 @@ def pang_one_hot_encode(seq: str) -> np.ndarray:
     return IN_MAP[seq_array.astype('int8')]
 
 
-def one_hot_encode(seq: str) -> np.ndarray:
-    """One-hot encode DNA sequence for SpliceAI model."""
-    if not isinstance(seq, str):
-        raise TypeError(f"Expected string, got {type(seq).__name__}")
-
-    valid_chars = set('ACGTN')
-    if not all(c.upper() in valid_chars for c in seq):
-        raise ValueError("Sequence contains invalid characters")
-
-    encoding_map = np.asarray([[0, 0, 0, 0],  # N
-                               [1, 0, 0, 0],  # A
-                               [0, 1, 0, 0],  # C
-                               [0, 0, 1, 0],  # G
-                               [0, 0, 0, 1]]) # T
-
-    seq = seq.upper().replace('A', '\x01').replace('C', '\x02')
-    seq = seq.replace('G', '\x03').replace('T', '\x04').replace('N', '\x00')
-
-    return encoding_map[np.frombuffer(seq.encode('latin1'), np.int8) % 5]
 
 
 def pangolin_predict_probs(seq: str, models: list = None) -> Tuple[List[float], List[float]]:
@@ -208,30 +146,30 @@ def pangolin_predict_probs(seq: str, models: list = None) -> Tuple[List[float],
     return donor_probs, acceptor_probs
 
 
-def sai_predict_probs(seq: str, models: list = None) -> Tuple[np.ndarray, np.ndarray]:
-    """Predict donor and acceptor probabilities using SpliceAI."""
-    if models is None:
-        models = _load_spliceai_models()
+def sai_predict_probs(seq: str) -> Tuple[np.ndarray, np.ndarray]:
+    """Predict acceptor and donor probabilities using OpenSpliceAI.
 
-    if not models:
-        raise ValueError("No SpliceAI models loaded")
+    Uses the OpenSpliceAI predict() function which handles encoding,
+    windowing, ensemble averaging, and softmax internally.
 
-    if len(seq) < 1000:
-        raise ValueError(f"Sequence too short: {len(seq)} (expected >= 1000)")
+    Returns (acceptor_probs, donor_probs) as numpy arrays matching the
+    full input sequence length.
+    """
+    from openspliceai.predict.predict import predict
+    import io, sys
 
-    x = one_hot_encode(seq)[None, :].astype(np.float32)
+    model_dir = _get_openspliceai_model_dir()
 
-    # Use direct model call instead of .predict() to avoid Jupyter kernel issues
-    preds = []
-    for model in models:
-        pred = model(x, training=False)
-        if hasattr(pred, 'numpy'):
-            pred = pred.numpy()
-        preds.append(pred)
-
-    y = np.mean(preds, axis=0)
-    y = y[0, :, 1:].T
-    return y[0, :], y[1, :]
+    # Suppress OpenSpliceAI's verbose print output
+    _stdout = sys.stdout
+    sys.stdout = io.StringIO()
+    try:
+        y = predict(seq, model_dir, flanking_size=10000)  # (seq_len, 3)
+    finally:
+        sys.stdout = _stdout
+
+    y = y.numpy()
+    return y[:, 1], y[:, 2]  # acceptor, donor
 
 
 def run_spliceai_seq(

geney/oncosplice.py

@@ -388,9 +388,10 @@ class Oncosplice:
         analysis_dict = {
             'reference_protein': self.reference_protein,
             'variant_protein': self.variant_protein,
+            'aligned_reference_protein': self.alignment.seqA,
+            'aligned_variant_protein': self.alignment.seqB,
             'reference_length': len(self.reference_protein),
             'variant_length': len(self.variant_protein),
-            # 'alignment_length': len(self.alignment.seqA),
             'oncosplice_score': self.score,
             'percentile': self.percentile,
             'number_of_deletions': len(self.deletions),

geney/pipelines.py

@@ -34,6 +34,10 @@ def oncosplice_pipeline(
         .generate_protein()
     )
 
+    # Truncate protein at first stop codon
+    if '*' in reference_transcript.protein:
+        reference_transcript.protein = reference_transcript.protein[:reference_transcript.protein.index('*') + 1]
+
     tl = TranscriptLibrary(reference_transcript, m)
     central_pos = m.central_position
 
@@ -71,6 +75,7 @@ def oncosplice_pipeline(
                 pd.Series({
                     "reference_mrna": reference_transcript.mature_mrna.seq,
                     "variant_mrna": variant_transcript.mature_mrna.seq,
+                    "conservation_vector": onco.conservation_vector,
                 }),
                 onco.get_analysis_series(),
             ])
@@ -123,6 +128,10 @@ def max_splicing_delta(
         .generate_protein()
     )
 
+    # Truncate protein at first stop codon
+    if '*' in reference_transcript.protein:
+        reference_transcript.protein = reference_transcript.protein[:reference_transcript.protein.index('*') + 1]
+
     tl = TranscriptLibrary(reference_transcript, m)
     splicing_results = tl.predict_splicing(
         m.central_position, engine=splicing_engine, inplace=True

geney/splice_graph.py

@@ -422,6 +422,10 @@ class SpliceSimulator:
             t.path_hash = _short_hash(tuple(donors + acceptors))
             t.generate_mature_mrna().generate_protein()
 
+            # Truncate protein at first stop codon
+            if '*' in t.protein:
+                t.protein = t.protein[:t.protein.index('*') + 1]
+
             # Filter out implausible IR paths (where cryptic sites compensate)
             if self._is_implausible_ir_path(t):
                 continue

{geney-1.4.41.dist-info → geney-1.4.44.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: geney
-Version: 1.4.41
+Version: 1.4.44
 Summary: A Python package for gene expression modeling.
 Home-page: https://github.com/nicolaslynn/geney
 Author: Nicolas Lynn
@@ -18,11 +18,9 @@ Requires-Dist: pandas==2.1.4
 Requires-Dist: biopython>=1.81
 Requires-Dist: matplotlib
 Requires-Dist: seaborn
-Requires-Dist: tensorflow>=2.8.0
-Requires-Dist: keras>=2.8.0
 Requires-Dist: torch
+Requires-Dist: openspliceai
 Requires-Dist: seqmat
-Requires-Dist: h5py
 Dynamic: author
 Dynamic: author-email
 Dynamic: classifier

geney-1.4.44.dist-info/RECORD

@@ -0,0 +1,16 @@
+geney/__init__.py,sha256=AIKhk1FPsnObK-GrBXU1T780KX0i1K-q34r4zh3ojok,993
+geney/engines.py,sha256=QfwzcLZvRPNVuiO7VQuSf-5Ay-4lYxw7J5zYygCPPeM,12288
+geney/oncosplice.py,sha256=rEVNhHugtzOvwicdjzRqbcyWG-KM0JIgvQa9Mpo23p4,18076
+geney/pipelines.py,sha256=ply3zS5zgA-4I-vCRF0_bJmZ4CtIrzSSpUQOsh3AZmo,4469
+geney/splice_graph.py,sha256=BnJDSOq3mePC0I0cFyfXyUyePANwhrk3LlBPoV1vzSs,24081
+geney/transcripts.py,sha256=BBgyeqF4jeIiHaD_bXxgOTXz19kdUgjcPVo4ClpcSUg,2594
+geney/variants.py,sha256=vjbiBH-duZ4TJZyXwXbQ_VmJxCFafjeDwLNTZg3ubSc,11832
+geney/models/openspliceai-mane/10000nt/model_10000nt_rs10.pt,sha256=ew1q9iOeJzkiJrutoxDunizgztFJ-2_f-JlsEsujzIU,2878124
+geney/models/openspliceai-mane/10000nt/model_10000nt_rs11.pt,sha256=OP5Eae_wn_5cqikbwiQkjIHtaKzKC3Ka-4dFCQmoEw4,2878124
+geney/models/openspliceai-mane/10000nt/model_10000nt_rs12.pt,sha256=SfOOP9Doe4HOItmG0RaOubLFVmi_1qqgZNoQgu5mhmc,2878124
+geney/models/openspliceai-mane/10000nt/model_10000nt_rs13.pt,sha256=uuJ1a0UvMdRbdGBIFlYdp5KE15G-N_RnSRpoNCyBmyA,2878124
+geney/models/openspliceai-mane/10000nt/model_10000nt_rs14.pt,sha256=z-J0ExHtAtqzBCSSpb4UApsY1jvmk3UelTQxWp5MYnE,2878124
+geney-1.4.44.dist-info/METADATA,sha256=pUGCeXBltgnICOYVdgH81g_sDo2LlPL486-4vYGLCrM,919
+geney-1.4.44.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+geney-1.4.44.dist-info/top_level.txt,sha256=O-FuNUMb5fn9dhZ-dYCgF0aZtfi1EslMstnzhc5IIVo,6
+geney-1.4.44.dist-info/RECORD,,

{geney-1.4.41.dist-info → geney-1.4.44.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (80.9.0)
+Generator: setuptools (80.10.2)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

geney-1.4.41.dist-info/RECORD

@@ -1,11 +0,0 @@
-geney/__init__.py,sha256=nkhniqCNWJzrb7xHgTDFEXSvRVdggb9ZCJ7ih7HEYq8,966
-geney/engines.py,sha256=9_oNsoluJsjdLC3cyWttjHF3cuQoy65FWgS4r7ehzek,14296
-geney/oncosplice.py,sha256=eGQQl9ftmoFENMYBWoJtenKWmzyxR9N1of5cZst_bHQ,18014
-geney/pipelines.py,sha256=gsy-gmHIi260SC5MKQ9IBSE0wko8Tvd7IC3wj083mPQ,3996
-geney/splice_graph.py,sha256=PANtLUAQiz578NZwxVlTSgboetnToHnQSkYpT0zbi_w,23931
-geney/transcripts.py,sha256=BBgyeqF4jeIiHaD_bXxgOTXz19kdUgjcPVo4ClpcSUg,2594
-geney/variants.py,sha256=vjbiBH-duZ4TJZyXwXbQ_VmJxCFafjeDwLNTZg3ubSc,11832
-geney-1.4.41.dist-info/METADATA,sha256=zuzWKIEeHSaFr08eRUjq3ZSiloOepcCD_QRG5ifS8j0,972
-geney-1.4.41.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-geney-1.4.41.dist-info/top_level.txt,sha256=O-FuNUMb5fn9dhZ-dYCgF0aZtfi1EslMstnzhc5IIVo,6
-geney-1.4.41.dist-info/RECORD,,