Bgen

README.md

@@ -41,7 +41,7 @@ To run PrediXcan Associations and MulTiXcan, you also need:
 * [h5py-cache (>=1.0.0)](https://pypi.python.org/pypi/h5py-cache/1.0) *Now folded into h5py
 
 To run prediction of biological mechanisms on individual-level data, you will also need:
-* [bgen_reader (>=3.0.3)](https://pypi.org/project/bgen-reader/)
+* [bgen (>=1.9.10)](https://pypi.org/project/bgen/)
 * [cyvcf2 (>=0.8.0)](https://brentp.github.io/cyvcf2)
 
 [R](https://www.r-project.org/) with [ggplot](http://ggplot2.org/) and [dplyr](https://cran.r-project.org/web/packages/dplyr/index.html) 

software/ToHDF5.py

@@ -4,7 +4,7 @@
 from metax import Logging, Utilities
 
 import numpy as np
-import h5py_cache
+import h5py
 import pandas
 
 def run(args):
@@ -18,7 +18,7 @@ def run(args):
     data = pandas.read_table(args.input)
 
     logging.info("Opening output")
-    f = h5py_cache.File(args.output, 'w', chunk_cache_mem_size=int(50 * (1024 ** 2)))
+    f = h5py.File(args.output, 'w', rdcc_nbytes=int(50 * (1024 ** 2)))
 
     n_genes = data.shape[1]-2
     n_samples = data.shape[0]
@@ -58,4 +58,4 @@ def run(args):
 
     args = parser.parse_args()
     Logging.configureLogging(args.verbosity)
-    run(args)
+    run(args)

software/conda_env.yaml

@@ -7,14 +7,12 @@ channels:
 dependencies:
   - python=3.9
   - pandas=2.2
-  - scipy=1.12
-  - numpy=1.26
   - pyliftover=0.4
   - statsmodels=0.14
   - h5py=3.12
-  - pyarrow=19.0
   - pip=25.0
   - pip:
     - cyvcf2==0.30.28
-    - bgen-reader==4.0.8
-
+    - bgen==1.9.10
+    - scipy==1.12
+    - numpy==1.26

software/metax/MainScreen.py

@@ -349,7 +349,7 @@ def __init__(self,args):
             def run(self):
                 try:
                     MetaXcan.run(self.args)
-                except Exceptions.ReportableException, e:
+                except Exceptions.ReportableException as e:
                     logging.error(e.msg)
                 except Exception as e:
                     logging.info("Exception when running task: %s", str(e))
@@ -368,12 +368,12 @@ def run(self):
                         work = self.works[i]
                         work.run()
                         del self.works[i]
-                except Exceptions.ReportableException, e:
+                except Exceptions.ReportableException as e:
                     logging.error(e.msg)
                 except Exception as e:
                     logging.info("Exception when running task: %s", str(e))
                 finally:
                     pass
 
         work = MetaxcanWorkWrapper(beta_args)
-        return work
+        return work

software/metax/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.7.5"
+__version__ = "0.8.2"
 
 def exitIf(doExit, Exception, msg):
     if doExit:

software/metax/genotype/BGENGenotype.py

@@ -1,25 +1,24 @@
-import bgen_reader
 import numpy
 import pandas
 import logging
+from bgen import BgenReader
 
 from ..misc import Genomics
 
 def bgen_file_geno_lines(file, variant_mapping = None, force_colon = False, use_rsid=False, whitelist=None, skip_palindromic=False, liftover_conversion=None):
     logging.log(9, "Processing bgen %s", file)
-    bgen = bgen_reader.read_bgen(file)
-    variants = bgen["variants"]
+    bfile = BgenReader(file)
     dict_mapping = variant_mapping is not None and type(variant_mapping) == dict
-    for variant in variants.itertuples():
+    for variant in bfile:
         if use_rsid:
             varid = variant.rsid
         else:
-            varid = variant.id
+            varid = variant.varid
 
         if force_colon:
             varid = varid.replace("_", ":")
 
-        alleles = variant.allele_ids.split(",")
+        alleles = variant.alleles
         if len(alleles) > 2:
             logging.info("variant %s is multiallelic, skipping", varid)
             continue
@@ -51,11 +50,10 @@ def bgen_file_geno_lines(file, variant_mapping = None, force_colon = False, use_
         # the alleles in the genotype might be swapped respect the variant in the mapping
         # You should verify if you must match it
 
-        v = bgen["genotype"][variant.Index].compute()
-        if v["phased"]:
-            d = numpy.apply_along_axis(lambda x: x[1] + x[3], 1, numpy.array(v["probs"], dtype=float))
+        if variant.is_phased:
+            d = numpy.apply_along_axis(lambda x: x[1] + x[3], 1, variant.probabilities)
         else:
-            d = numpy.apply_along_axis(lambda x: x[1] + x[2] * 2, 1, numpy.array(v["probs"], dtype=float))
+            d = numpy.apply_along_axis(lambda x: x[1] + x[2] * 2, 1, variant.probabilities)
 
         #e = bgen_reader.allele_expectation(bgen, variant.Index)
         #d2 = bgen_reader.compute_dosage(e, alt=1)
@@ -70,7 +68,6 @@ def bgen_files_geno_lines(files, variant_mapping = None, force_colon = False, us
 
 def get_samples(path):
     logging.info("Opening bgen to get samples")
-    bgen = bgen_reader.read_bgen(path, verbose=False)
-    samples = bgen["samples"].values
-    samples = pandas.DataFrame({"FID":samples, "IID":samples})[["FID", "IID"]]
+    bfile = BgenReader(path)
+    samples = pandas.DataFrame({"FID":bfile.samples, "IID":bfile.samples})[["FID", "IID"]]
     return samples

software/setup.py

@@ -29,14 +29,23 @@ def read(fname):
                             'MulTiXcan.py',
                             'SMulTiXcan.py'],
                  description=["TBD"],
-                 install_requires=['scipy>=1.2.2', 'numpy>=1.14.2', 'pandas>=0.22.0', 'patsy>=0.5.0',
-                                   'statsmodels>=0.10.0', 'h5py>=2.7.1', 'h5py-cache>=1.0', 'bgen_reader>=3.0.3', 'cyvcf2>=0.8.0'],
+                 install_requires=[
+                     'numpy>=1.23,<2',
+                     'scipy>=1.10,<2',
+                     'pandas>=2.0,<3',
+                     'patsy>=0.5.6',
+                     'statsmodels>=0.14,<0.15',
+                     'h5py>=3.10,<4',
+                     'pyliftover>=0.4',
+                     'bgen==1.9.10',
+                     'cyvcf2>=0.30',
+                 ],
                  extras_require={"test": ["sqlalchemy"]},
                  long_description=read('Readme.md'),
                  keywords=['TBD'],
                  test_suite='tests',
                  package_data={'tests/files':['*']},
-                 python_requires='>=3.5',
+                 python_requires='>=3.9',
                  classifiers=[
                     "Development Status :: 4 - Beta",
                     "Topic :: Utilities",