exprmat 0.1.3__tar.gz → 0.1.4__tar.gz

{exprmat-0.1.3 → exprmat-0.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.2
 Name: exprmat
-Version: 0.1.3
+Version: 0.1.4
 Summary: routines to process expression matrices
 Home-page: https://github.com/yang-z-03/exprmat
 Author: Zheng Yang

{exprmat-0.1.3 → exprmat-0.1.4}/exprmat/preprocessing/__init__.py

@@ -104,11 +104,11 @@ def highly_variable(
         adata.var[dest + '.dispersions.norm'] = df['norm.disp']
         if use_bounds:
             adata.var[dest + '.hvg'] = (
-                df['means'] > min_mean & 
-                df['means'] < max_mean & 
-                df['norm.disp'] > min_disp &
-                df['norm.disp'] < max_disp
-            )
+                (df['means'] > min_mean) & 
+                (df['means'] < max_mean) & 
+                (df['norm.disp'] > min_disp) &
+                (df['norm.disp'] < max_disp)
+            ).tolist()
 
         else:
             if n_top_genes > adata.n_vars:
@@ -129,11 +129,11 @@ def highly_variable(
         adata.var[dest + '.dispersions.norm'] = df['norm.disp']
         if use_bounds:
             adata.var[dest + '.hvg'] = (
-                df['means'] > min_mean & 
-                df['means'] < max_mean & 
-                df['norm.disp'] > min_disp &
-                df['norm.disp'] < max_disp
-            )
+                (df['means'] > min_mean) & 
+                (df['means'] < max_mean) & 
+                (df['norm.disp'] > min_disp) &
+                (df['norm.disp'] < max_disp)
+            ).tolist()
 
         else:
             if n_top_genes > adata.n_vars:

{exprmat-0.1.3 → exprmat-0.1.4}/exprmat/preprocessing/filter.py

@@ -449,7 +449,7 @@ def highly_variable_nb(
 
     # actually do the normalization
     df["norm.disp"] = (df["dispersions"] - disp_stat["avg"]) / disp_stat["dev"]
-    df.index = var.index
+    # df.index = var.index
     return df
 
 
@@ -477,7 +477,7 @@ def highly_variable_nb_log(
 
     # actually do the normalization
     df["norm.disp"] = (df["dispersions"] - disp_stat["avg"]) / disp_stat["dev"]
-    df.index = var.index
+    # df.index = var.index
     return df
 
 

{exprmat-0.1.3 → exprmat-0.1.4}/exprmat/reader/experiment.py

@@ -24,7 +24,7 @@ import os
 
 from exprmat.reader.metadata import metadata, load_metadata
 from exprmat.data.finders import get_genome
-from exprmat.reader.matcher import read_mtx_rna
+from exprmat.reader.matcher import read_mtx_rna, read_h5ad_rna, read_table_rna
 from exprmat.ansi import warning, info, error, red, green
 
 
@@ -67,10 +67,31 @@ class experiment:
 
             if i_mod == 'rna':
                 if not 'rna' in self.modalities.keys(): self.modalities['rna'] = {}
-                self.modalities['rna'][i_sample] = read_mtx_rna(
-                    src = i_loc, prefix = '', metadata = meta, sample = i_sample,
-                    raw = False, default_taxa = i_taxa
-                )
+
+                # we automatically infer from the given location names to select
+                # the correct way of loading samples:
+
+                if i_loc.endswith('.tsv.gz') or i_loc.endswith('.csv.gz') or \
+                    i_loc.endswith('.tsv') or i_loc.endswith('.csv'):
+
+                    self.modalities['rna'][i_sample] = read_table_rna(
+                        src = i_loc, metadata = meta, sample = i_sample,
+                        raw = False, default_taxa = i_taxa
+                    )
+
+                elif i_loc.endswith('.h5') or i_loc.endswith('.h5ad'):
+
+                    self.modalities['rna'][i_sample] = read_h5ad_rna(
+                        src = i_loc, metadata = meta, sample = i_sample,
+                        raw = False, default_taxa = i_taxa
+                    )
+
+                else:
+                    self.modalities['rna'][i_sample] = read_mtx_rna(
+                        src = i_loc, prefix = '', metadata = meta, sample = i_sample,
+                        raw = False, default_taxa = i_taxa
+                    )
+                    
                 self.modalities['rna'][i_sample].var = \
                     experiment.search_genes(self.modalities['rna'][i_sample].var_names.tolist())
             
@@ -612,6 +633,7 @@ class experiment:
         orig = adata.obs[cluster_key].tolist()
         merge = [x if x not in clusters else y.replace(',', '.') for x, y in zip(orig, temp)]
         del adata.obs['.leiden.temp']
+        del adata.uns['.leiden.temp']
         adata.obs[key_added] = merge
 
         # for all categorical types:

{exprmat-0.1.3 → exprmat-0.1.4}/exprmat/reader/matcher.py

@@ -230,9 +230,116 @@ def read_mtx_rna(
     return final
 
 
+def read_table_rna(
+        src: str, 
+        metadata: metadata, sample: str, raw: bool = False,
+        default_taxa = 'mmu'
+    ):
+    '''
+    Read from compressed or plain text tables.
+
+    Parameters
+    ----------
+
+    src : str
+        The directory to the matrix.
+    
+    metadata : exprmat.reader.metadata.metadata
+        The sample specific metadata table.
+    
+    sample : str
+        The sample name according to the metadata.
+
+    raw : bool
+        Whether this expression matrix is confirmed to be unfiltered one. Unfiltered expression
+        matrix contains all detected beads or droplets, including the empty drops and doublets
+        at higher probabilities. We guess this field by default to see if a sample contains
+        unexpectedly high number of cells.
+    '''
+
+    # try:
+    
+    # in csv/tsv texts, genes are rows and cells are columns.
+    adata = sc.read_text(
+        src, delimiter = '\t', first_column_names = True, dtype = 'float32'
+    ).T
+
+    # except Exception as ex: 
+    #     warning('error occurs when reading matrix files:')
+    #     warning(str(ex))
+    #     return None
+
+    final = match_matrix_rna(
+        adata, metadata, sample, 
+        force_filter = raw, default_taxa = default_taxa
+    )
+    
+    del adata
+    return final
+
+
+def read_h5ad_rna(
+        src: str, 
+        metadata: metadata, sample: str, raw: bool = False,
+        default_taxa = 'mmu'
+    ):
+    '''
+    Read from a given h5ad. This typically occurs when you have to prepare the data yourself
+    from others who give you a complex atlas. The assignment of samples and batches happens
+    within a single h5ad file. Then, you would like to keep nearly all of the metadata you
+    prepared by hand as it is. Only `modality`, `taxa` and `ubc` are generated for you.
+
+    Thus, you must manually supply `sample`, `batch` and `group` yourself. Otherwise the program
+    may not work as usual. It permits multiple samples in one h5ad file.
+
+    Parameters
+    ----------
+
+    src : str
+        The directory to the matrix.
+    
+    prefix : str
+        Prefix before the file name convention.
+    
+    metadata : exprmat.reader.metadata.metadata
+        The sample specific metadata table.
+    
+    sample : str
+        The sample name according to the metadata.
+
+    raw : bool
+        Whether this expression matrix is confirmed to be unfiltered one. Unfiltered expression
+        matrix contains all detected beads or droplets, including the empty drops and doublets
+        at higher probabilities. We guess this field by default to see if a sample contains
+        unexpectedly high number of cells.
+    '''
+
+    # try:
+    
+    adata = sc.read_h5ad(src)
+
+    # except Exception as ex: 
+    #     warning('error occurs when reading matrix files:')
+    #     warning(str(ex))
+    #     return None
+
+    final = match_matrix_rna(
+        adata, metadata, sample, 
+        force_filter = raw, default_taxa = default_taxa,
+        # this switch tells the program not to alter the cell names (keep them
+        # directly to `barcodes`, not adding prefix to them), and keep the
+        # sample, batch, and group columns as is. 
+        do_not_alter_obs_names = True
+    )
+    
+    del adata
+    return final
+
+
 def match_matrix_rna(
         adata, metadata: metadata, sample: str, 
-        force_filter = False, default_taxa = 'mmu'
+        force_filter = False, default_taxa = 'mmu',
+        do_not_alter_obs_names = False
     ):
 
     # if more than 50000 cells in a single matrix, we just believe that it is
@@ -249,8 +356,9 @@ def match_matrix_rna(
     assert len(rows) == 1
     props = rows.iloc[0]
 
-    # append the sample name to the barcode
-    adata_f.obs_names = props['sample'] + ':' + adata_f.obs_names
+    if not do_not_alter_obs_names:
+        # append the sample name to the barcode
+        adata_f.obs_names = props['sample'] + ':' + adata_f.obs_names
 
     # map gene naming
     gname = adata_f.var_names.tolist()
@@ -320,6 +428,8 @@ def match_matrix_rna(
     # attach cell metadata onto the obs slot.
     
     for k in props.index.tolist():
+        if do_not_alter_obs_names:
+            if k in ['sample', 'batch', 'group']: continue
         if k not in ['location']: final.obs[k] = props[k]
     
     final.obs['barcode'] = final.obs_names

{exprmat-0.1.3 → exprmat-0.1.4}/exprmat.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.2
 Name: exprmat
-Version: 0.1.3
+Version: 0.1.4
 Summary: routines to process expression matrices
 Home-page: https://github.com/yang-z-03/exprmat
 Author: Zheng Yang

{exprmat-0.1.3 → exprmat-0.1.4}/exprmat.egg-info/SOURCES.txt

@@ -9,7 +9,6 @@ exprmat.egg-info/dependency_links.txt
 exprmat.egg-info/not-zip-safe
 exprmat.egg-info/requires.txt
 exprmat.egg-info/top_level.txt
-exprmat.egg-info/.ipynb_checkpoints/requires-checkpoint.txt
 exprmat/clustering/__init__.py
 exprmat/clustering/leiden.py
 exprmat/clustering/spectral.py
@@ -24,6 +23,7 @@ exprmat/data/gff.py
 exprmat/data/private.py
 exprmat/data/ranges.py
 exprmat/data/hsa/genome.tsv.gz
+exprmat/data/mmu/gencode.tsv.gz
 exprmat/data/mmu/genome.tsv.gz
 exprmat/descriptive/__init__.py
 exprmat/descriptive/de.py

{exprmat-0.1.3 → exprmat-0.1.4}/setup.py

@@ -3,7 +3,7 @@ from setuptools import setup, find_packages
 
 setup(
     name                 = 'exprmat',
-    version              = '0.1.3',
+    version              = '0.1.4',
     description          = 'routines to process expression matrices',
     url                  = 'https://github.com/yang-z-03/exprmat',
     author               = 'Zheng Yang',
@@ -34,9 +34,12 @@ setup(
         'pynndescent',
         'gseapy'
     ],
-    include_package_data = True,
+    include_package_data = False,
     package_data         = {
-        'exprmat.data': ['mmu/*.gz', 'hsa/*.gz']
+        'exprmat.data': [
+            'mmu/*.tsv.gz', 
+            'hsa/*.tsv.gz'
+        ]
     },
     zip_safe             = False
 )

exprmat-0.1.3/exprmat.egg-info/.ipynb_checkpoints/requires-checkpoint.txt

@@ -1,22 +0,0 @@
-umap-learn>=0.5.0
-scvi-tools>=1.3.0
-scanpy
-anndata
-mudata
-pandas
-numpy
-matplotlib
-seaborn
-torch>=2.0
-annoy
-scikit-image
-scikit-misc
-igraph
-bbknn
-scanorama
-harmonypy
-pymde
-datashader
-fa2_modified
-pynndescent
-gseapy