gsrap 0.7.1__tar.gz → 0.7.2__tar.gz

{gsrap-0.7.1 → gsrap-0.7.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: gsrap
-Version: 0.7.1
+Version: 0.7.2
 Summary: 
 License: GNU General Public License v3.0
 Author: Gioele Lazzari
@@ -17,9 +17,11 @@ Requires-Dist: cobra (>=0.29)
 Requires-Dist: colorlog (>=6.9.0)
 Requires-Dist: gdown (>=5.2.0)
 Requires-Dist: gempipe (>=1.38.1)
+Requires-Dist: matplotlib (>=3.9.0)
 Requires-Dist: memote (>=0.17.0)
 Requires-Dist: openpyxl (>=3.1.0)
 Requires-Dist: pandas (>=2.0.0)
+Requires-Dist: xlsxwriter (>=3.1.0)
 Description-Content-Type: text/markdown
 
 Source code for `gsrap`.

{gsrap-0.7.1 → gsrap-0.7.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "gsrap"
-version = "0.7.1"
+version = "0.7.2"
 description = ""
 authors = ["Gioele Lazzari"]
 license = "GNU General Public License v3.0"
@@ -16,6 +16,9 @@ gempipe = ">=1.38.1"
 gdown = ">=5.2.0"
 colorlog = ">=6.9.0"
 memote = ">=0.17.0"
+matplotlib = ">=3.9.0"
+xlsxwriter = ">=3.1.0"
+
 
 [build-system]
 requires = ["poetry-core>=1.0.0"]

{gsrap-0.7.1 → gsrap-0.7.2}/src/gsrap/.ipynb_checkpoints/__init__-checkpoint.py

@@ -72,9 +72,9 @@ def main():
     parsedb_parser.add_argument("--precursors", action='store_true', help="Verify biosynthesis of biomass precursors and show blocked ones.")
     parsedb_parser.add_argument("--biosynth", action='store_true', help="Check biosynthesis of all metabolites and detect dead-ends.")
     parsedb_parser.add_argument("-e", "--eggnog", nargs='+', metavar='', type=str, default='-', help="Path to the optional eggnog-mapper annotation table(s).")
-    #parsedb_parser.add_argument("-z", "--zeroes", action='store_true', help="Show maps/modules with 0%% coverage, in addition to partials (use only with --progress).")
     parsedb_parser.add_argument("--goodbefore", metavar='', type=str, default='-', help="Syntax is {pure_mid}-{rid1}-{rid2}. From top to bottom, build the universe until reaction {rid1}, transport {rid2} and metabolite {pure_mid} are reached.")
     parsedb_parser.add_argument("--onlyauthor", metavar='', type=str, default='-', help="Build the universe by parsing contents of the specified author ID only. Contents affected by --goodbefore are parsed anyway.")
+    parsedb_parser.add_argument("--nofigs", action='store_true', help="Do not generate figures.")
     
     
     # add arguments for the 'mkmodel' command
@@ -94,6 +94,7 @@ def main():
     mkmodel_parser.add_argument("--conditional", metavar='', type=float, default=0.5, help="Expected minimum fraction of reactions in a biosynthetic pathway for an actually present conditional biomass precursor.")
     mkmodel_parser.add_argument("--biosynth", action='store_true', help="Check biosynthesis of all metabolites and detect dead-ends.")
     mkmodel_parser.add_argument("-b", "--biomass", metavar='', type=str, default='-', help="Strain ID associated to experimental biomass data.")
+    mkmodel_parser.add_argument("--nofigs", action='store_true', help="Do not generate figures.")
     
     
     # add arguments for the 'runsims' command
@@ -110,6 +111,7 @@ def main():
     runsims_parser.add_argument("--omission", action='store_true', help="Perform single omission experiments to study auxotrophies.")
     runsims_parser.add_argument("--essential", action='store_true', help="Predict essential genes (single-gene knock-out simulations).")
     runsims_parser.add_argument("--factors", action='store_true', help="Predict putative growth factors.")
+    runsims_parser.add_argument("--nofigs", action='store_true', help="Do not generate figures.")
 
     
     # check the inputted subcommand, automatic sys.exit(1) if a bad subprogram was specied. 

{gsrap-0.7.1 → gsrap-0.7.2}/src/gsrap/__init__.py

@@ -72,9 +72,9 @@ def main():
     parsedb_parser.add_argument("--precursors", action='store_true', help="Verify biosynthesis of biomass precursors and show blocked ones.")
     parsedb_parser.add_argument("--biosynth", action='store_true', help="Check biosynthesis of all metabolites and detect dead-ends.")
     parsedb_parser.add_argument("-e", "--eggnog", nargs='+', metavar='', type=str, default='-', help="Path to the optional eggnog-mapper annotation table(s).")
-    #parsedb_parser.add_argument("-z", "--zeroes", action='store_true', help="Show maps/modules with 0%% coverage, in addition to partials (use only with --progress).")
     parsedb_parser.add_argument("--goodbefore", metavar='', type=str, default='-', help="Syntax is {pure_mid}-{rid1}-{rid2}. From top to bottom, build the universe until reaction {rid1}, transport {rid2} and metabolite {pure_mid} are reached.")
     parsedb_parser.add_argument("--onlyauthor", metavar='', type=str, default='-', help="Build the universe by parsing contents of the specified author ID only. Contents affected by --goodbefore are parsed anyway.")
+    parsedb_parser.add_argument("--nofigs", action='store_true', help="Do not generate figures.")
     
     
     # add arguments for the 'mkmodel' command
@@ -94,6 +94,7 @@ def main():
     mkmodel_parser.add_argument("--conditional", metavar='', type=float, default=0.5, help="Expected minimum fraction of reactions in a biosynthetic pathway for an actually present conditional biomass precursor.")
     mkmodel_parser.add_argument("--biosynth", action='store_true', help="Check biosynthesis of all metabolites and detect dead-ends.")
     mkmodel_parser.add_argument("-b", "--biomass", metavar='', type=str, default='-', help="Strain ID associated to experimental biomass data.")
+    mkmodel_parser.add_argument("--nofigs", action='store_true', help="Do not generate figures.")
     
     
     # add arguments for the 'runsims' command
@@ -110,6 +111,7 @@ def main():
     runsims_parser.add_argument("--omission", action='store_true', help="Perform single omission experiments to study auxotrophies.")
     runsims_parser.add_argument("--essential", action='store_true', help="Predict essential genes (single-gene knock-out simulations).")
     runsims_parser.add_argument("--factors", action='store_true', help="Predict putative growth factors.")
+    runsims_parser.add_argument("--nofigs", action='store_true', help="Do not generate figures.")
 
     
     # check the inputted subcommand, automatic sys.exit(1) if a bad subprogram was specied. 

{gsrap-0.7.1 → gsrap-0.7.2}/src/gsrap/commons/.ipynb_checkpoints/excelhub-checkpoint.py

@@ -1,8 +1,20 @@
 import pandas as pnd
 
 
+from .figures import figure_df_C_F1
 
-def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
+
+
+def write_excel_model(model, filepath, nofigs, df_E, df_B, df_P, df_S, df_C=None):
+    
+    
+    # generate figures
+    if nofigs == False:
+              
+        if df_C is not None:
+            df_C_F1 = figure_df_C_F1(df_C)
+        
+        
     
     df_M = []
     df_R = []
@@ -33,6 +45,12 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
         df_S.insert(0, 'mid', '')  # new columns as first
         df_S['mid'] = df_S.index
         df_S = df_S.reset_index(drop=True)
+        
+    # format df_C: universal reaction coverage
+    if df_C is not None:
+        df_C.insert(0, 'kr', '')  # new columns as first
+        df_C['kr'] = df_C.index
+        df_C = df_C.reset_index(drop=True)
             
         
     for m in model.metabolites: 
@@ -81,7 +99,7 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
     df_R = pnd.DataFrame.from_records(df_R)
     df_T = pnd.DataFrame.from_records(df_T)
     df_A = pnd.DataFrame.from_records(df_A)
-    with pnd.ExcelWriter(filepath) as writer:
+    with pnd.ExcelWriter(filepath, engine='xlsxwriter') as writer:
         df_M.to_excel(writer, sheet_name='Metabolites', index=False)
         df_R.to_excel(writer, sheet_name='Reactions', index=False)
         df_T.to_excel(writer, sheet_name='Transporters', index=False)
@@ -90,7 +108,12 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
         if df_B is not None: df_B.to_excel(writer, sheet_name='Biomass', index=False)
         if df_P is not None and len(df_P)!=0: df_P.to_excel(writer, sheet_name='Biolog®', index=False)
         if df_S is not None and len(df_S.columns)>2: df_S.to_excel(writer, sheet_name='Biosynth', index=False) 
-        
+        if df_C is not None: 
+            df_C.to_excel(writer, sheet_name='Coverage', index=False) 
+            if nofigs == False:
+                worksheet = writer.sheets['Coverage']
+                worksheet.insert_image('A1', 'df_C_F1.png', {'image_data': df_C_F1})
+            
         
     sheets_dict = {
         'model_id': model.id,
@@ -102,6 +125,7 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
         'Biomass': df_B,
         'Biolog': df_P,
         'Biosynth': df_S,
+        'Coverage': df_C,
     }
     return sheets_dict
 

gsrap-0.7.2/src/gsrap/commons/.ipynb_checkpoints/figures-checkpoint.py

@@ -0,0 +1,105 @@
+from io import BytesIO
+
+import numpy as np
+import pandas as pnd
+
+from scipy.spatial.distance import pdist
+from scipy.cluster.hierarchy import linkage, cut_tree, dendrogram, leaves_list
+
+import matplotlib.pyplot as plt
+from matplotlib.patches import Patch
+
+
+
+def figure_df_C_F1(df_coverage):
+
+    bin_matrix = df_coverage[[i for i in df_coverage.columns if i not in ['map_ids', 'modeled']]]
+    strains = bin_matrix.columns
+    bin_matrix = bin_matrix.T  # features in column
+
+    # pdist() / linkage() will loose the accession information. So here we save a dict: 
+    index_to_strain = {i: strain for i, strain in enumerate(bin_matrix.index)}
+
+    # Calculate the linkage matrix using Ward clustering and Jaccard dissimilarity
+    distances = pdist(bin_matrix, 'jaccard')
+    linkage_matrix = linkage(distances, method='ward')
+
+
+    # PART 0: create the frame
+    fig, axs = plt.subplots(
+        nrows=2, ncols=2, 
+        figsize=(15, 10), 
+        gridspec_kw={  # suplots width proportions. 
+            'width_ratios': [0.5, 1.0],
+            'height_ratios': [0.015, 0.985]
+        }
+    ) 
+
+    # PART 1: dendrogram
+    dn = dendrogram(
+        linkage_matrix, ax=axs[1,0],
+        orientation='left',
+        color_threshold=0, above_threshold_color='black',
+    )
+
+
+    ### PART 2: heatmap
+    ord_leaves = leaves_list(linkage_matrix)
+    ord_leaves = np.flip(ord_leaves)  # because leaves are returned in the inverse sense.
+    ord_leaves = [index_to_strain[i] for i in ord_leaves]  # convert index as number to index as accession
+    bin_matrix = bin_matrix.loc[ord_leaves, :]  # reordered dataframe.
+    axs[1,1].matshow(
+        bin_matrix,  
+        cmap='viridis',
+        aspect='auto', # non-squared pixels to fit the axis
+    )
+
+
+    ### PART 3: coverage bar
+    axs[0,1].matshow(
+        df_coverage[['modeled']].T,  
+        cmap='cool_r',
+        aspect='auto', # non-squared pixels to fit the axis
+    )
+
+
+    ### PART 4: legends
+    legend_feat = [
+        Patch(facecolor=plt.colormaps.get_cmap('viridis')(0.0), edgecolor='black', label='Absent'),
+        Patch(facecolor=plt.colormaps.get_cmap('viridis')(1.0), edgecolor='black', label='Probably present'),
+    ]
+    legend_cov = [
+        Patch(facecolor=plt.colormaps.get_cmap('cool_r')(0.0), edgecolor='black', label='Not modeled'),
+        Patch(facecolor=plt.colormaps.get_cmap('cool_r')(1.0), edgecolor='black', label='Modeled'),
+    ]
+    l1 = axs[1,0].legend(handles=legend_cov, title='Universe coverage', loc='upper left')
+    l2 = axs[1,0].legend(handles=legend_feat, title='KEGG reaction in strain', loc='lower left')
+    axs[1,0].add_artist(l1)  # keep both legends visible
+
+
+    ### PART 5: aesthetics
+    plt.subplots_adjust(wspace=0, hspace=0)  # adjust the space between subplots: 
+    axs[0,0].axis('off')  # remove frame and axis
+    axs[1,0].axis('off')  # remove frame and axis
+
+    axs[0,1].yaxis.set_visible(False)  # remove ticks, tick labels, axis label
+
+    axs[1,1].xaxis.set_ticks([])       # remove ticks
+    axs[1,1].set_xticklabels([])       # remove tick labels
+    axs[1,1].xaxis.set_label_position("bottom")
+    axs[1,1].set_xlabel("KEGG reactions")
+
+    axs[1,1].yaxis.set_ticks([])       # remove ticks
+    axs[1,1].set_yticklabels([])       # remove tick labels
+    axs[1,1].yaxis.set_label_position("right")
+    axs[1,1].set_ylabel(f"{len(strains)} strains", rotation=270, labelpad=13)  # labelpad is in points (1 point = 1/72 inch)
+
+
+    ### PART 6: save fig
+    buf = BytesIO()
+    fig.savefig(buf, dpi=300, bbox_inches='tight')  # labelpad is in inches (1 point = 1/72 inch)
+    plt.close(fig)
+    buf.seek(0)  # rewind the buffer to the beginning
+    
+    
+    return buf

{gsrap-0.7.1 → gsrap-0.7.2}/src/gsrap/commons/excelhub.py

@@ -1,8 +1,20 @@
 import pandas as pnd
 
 
+from .figures import figure_df_C_F1
 
-def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
+
+
+def write_excel_model(model, filepath, nofigs, df_E, df_B, df_P, df_S, df_C=None):
+    
+    
+    # generate figures
+    if nofigs == False:
+              
+        if df_C is not None:
+            df_C_F1 = figure_df_C_F1(df_C)
+        
+        
     
     df_M = []
     df_R = []
@@ -33,6 +45,12 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
         df_S.insert(0, 'mid', '')  # new columns as first
         df_S['mid'] = df_S.index
         df_S = df_S.reset_index(drop=True)
+        
+    # format df_C: universal reaction coverage
+    if df_C is not None:
+        df_C.insert(0, 'kr', '')  # new columns as first
+        df_C['kr'] = df_C.index
+        df_C = df_C.reset_index(drop=True)
             
         
     for m in model.metabolites: 
@@ -81,7 +99,7 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
     df_R = pnd.DataFrame.from_records(df_R)
     df_T = pnd.DataFrame.from_records(df_T)
     df_A = pnd.DataFrame.from_records(df_A)
-    with pnd.ExcelWriter(filepath) as writer:
+    with pnd.ExcelWriter(filepath, engine='xlsxwriter') as writer:
         df_M.to_excel(writer, sheet_name='Metabolites', index=False)
         df_R.to_excel(writer, sheet_name='Reactions', index=False)
         df_T.to_excel(writer, sheet_name='Transporters', index=False)
@@ -90,7 +108,12 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
         if df_B is not None: df_B.to_excel(writer, sheet_name='Biomass', index=False)
         if df_P is not None and len(df_P)!=0: df_P.to_excel(writer, sheet_name='Biolog®', index=False)
         if df_S is not None and len(df_S.columns)>2: df_S.to_excel(writer, sheet_name='Biosynth', index=False) 
-        
+        if df_C is not None: 
+            df_C.to_excel(writer, sheet_name='Coverage', index=False) 
+            if nofigs == False:
+                worksheet = writer.sheets['Coverage']
+                worksheet.insert_image('A1', 'df_C_F1.png', {'image_data': df_C_F1})
+            
         
     sheets_dict = {
         'model_id': model.id,
@@ -102,6 +125,7 @@ def write_excel_model(model, filepath, df_E, df_B, df_P, df_S):
         'Biomass': df_B,
         'Biolog': df_P,
         'Biosynth': df_S,
+        'Coverage': df_C,
     }
     return sheets_dict
 

gsrap-0.7.2/src/gsrap/commons/figures.py

@@ -0,0 +1,105 @@
+from io import BytesIO
+
+import numpy as np
+import pandas as pnd
+
+from scipy.spatial.distance import pdist
+from scipy.cluster.hierarchy import linkage, cut_tree, dendrogram, leaves_list
+
+import matplotlib.pyplot as plt
+from matplotlib.patches import Patch
+
+
+
+def figure_df_C_F1(df_coverage):
+
+    bin_matrix = df_coverage[[i for i in df_coverage.columns if i not in ['map_ids', 'modeled']]]
+    strains = bin_matrix.columns
+    bin_matrix = bin_matrix.T  # features in column
+
+    # pdist() / linkage() will loose the accession information. So here we save a dict: 
+    index_to_strain = {i: strain for i, strain in enumerate(bin_matrix.index)}
+
+    # Calculate the linkage matrix using Ward clustering and Jaccard dissimilarity
+    distances = pdist(bin_matrix, 'jaccard')
+    linkage_matrix = linkage(distances, method='ward')
+
+
+    # PART 0: create the frame
+    fig, axs = plt.subplots(
+        nrows=2, ncols=2, 
+        figsize=(15, 10), 
+        gridspec_kw={  # suplots width proportions. 
+            'width_ratios': [0.5, 1.0],
+            'height_ratios': [0.015, 0.985]
+        }
+    ) 
+
+    # PART 1: dendrogram
+    dn = dendrogram(
+        linkage_matrix, ax=axs[1,0],
+        orientation='left',
+        color_threshold=0, above_threshold_color='black',
+    )
+
+
+    ### PART 2: heatmap
+    ord_leaves = leaves_list(linkage_matrix)
+    ord_leaves = np.flip(ord_leaves)  # because leaves are returned in the inverse sense.
+    ord_leaves = [index_to_strain[i] for i in ord_leaves]  # convert index as number to index as accession
+    bin_matrix = bin_matrix.loc[ord_leaves, :]  # reordered dataframe.
+    axs[1,1].matshow(
+        bin_matrix,  
+        cmap='viridis',
+        aspect='auto', # non-squared pixels to fit the axis
+    )
+
+
+    ### PART 3: coverage bar
+    axs[0,1].matshow(
+        df_coverage[['modeled']].T,  
+        cmap='cool_r',
+        aspect='auto', # non-squared pixels to fit the axis
+    )
+
+
+    ### PART 4: legends
+    legend_feat = [
+        Patch(facecolor=plt.colormaps.get_cmap('viridis')(0.0), edgecolor='black', label='Absent'),
+        Patch(facecolor=plt.colormaps.get_cmap('viridis')(1.0), edgecolor='black', label='Probably present'),
+    ]
+    legend_cov = [
+        Patch(facecolor=plt.colormaps.get_cmap('cool_r')(0.0), edgecolor='black', label='Not modeled'),
+        Patch(facecolor=plt.colormaps.get_cmap('cool_r')(1.0), edgecolor='black', label='Modeled'),
+    ]
+    l1 = axs[1,0].legend(handles=legend_cov, title='Universe coverage', loc='upper left')
+    l2 = axs[1,0].legend(handles=legend_feat, title='KEGG reaction in strain', loc='lower left')
+    axs[1,0].add_artist(l1)  # keep both legends visible
+
+
+    ### PART 5: aesthetics
+    plt.subplots_adjust(wspace=0, hspace=0)  # adjust the space between subplots: 
+    axs[0,0].axis('off')  # remove frame and axis
+    axs[1,0].axis('off')  # remove frame and axis
+
+    axs[0,1].yaxis.set_visible(False)  # remove ticks, tick labels, axis label
+
+    axs[1,1].xaxis.set_ticks([])       # remove ticks
+    axs[1,1].set_xticklabels([])       # remove tick labels
+    axs[1,1].xaxis.set_label_position("bottom")
+    axs[1,1].set_xlabel("KEGG reactions")
+
+    axs[1,1].yaxis.set_ticks([])       # remove ticks
+    axs[1,1].set_yticklabels([])       # remove tick labels
+    axs[1,1].yaxis.set_label_position("right")
+    axs[1,1].set_ylabel(f"{len(strains)} strains", rotation=270, labelpad=13)  # labelpad is in points (1 point = 1/72 inch)
+
+
+    ### PART 6: save fig
+    buf = BytesIO()
+    fig.savefig(buf, dpi=300, bbox_inches='tight')  # labelpad is in inches (1 point = 1/72 inch)
+    plt.close(fig)
+    buf.seek(0)  # rewind the buffer to the beginning
+    
+    
+    return buf

{gsrap-0.7.1 → gsrap-0.7.2}/src/gsrap/mkmodel/.ipynb_checkpoints/mkmodel-checkpoint.py

@@ -141,7 +141,7 @@ def create_model_incore(params):
     cobra.io.write_sbml_model(model, f'{args.outdir}/{model.id}.xml')        # SBML   # groups are saved only to SBML
     logger.info(f"'{args.outdir}/{model.id}.xml' created!")
     force_id_on_sbml(f'{args.outdir}/{model.id}.xml', model.id)   # force introduction of the 'id=""' field
-    sheets_dict = write_excel_model(model, f'{args.outdir}/{model.id}.mkmodel.xlsx', None, df_B, df_P, df_S)  
+    sheets_dict = write_excel_model(model, f'{args.outdir}/{model.id}.mkmodel.xlsx', args.nofigs, None, df_B, df_P, df_S)  
     logger.info(f"'{args.outdir}/{model.id}.mkmodel.xlsx' created!")
         
         

{gsrap-0.7.1 → gsrap-0.7.2}/src/gsrap/mkmodel/mkmodel.py

@@ -141,7 +141,7 @@ def create_model_incore(params):
     cobra.io.write_sbml_model(model, f'{args.outdir}/{model.id}.xml')        # SBML   # groups are saved only to SBML
     logger.info(f"'{args.outdir}/{model.id}.xml' created!")
     force_id_on_sbml(f'{args.outdir}/{model.id}.xml', model.id)   # force introduction of the 'id=""' field
-    sheets_dict = write_excel_model(model, f'{args.outdir}/{model.id}.mkmodel.xlsx', None, df_B, df_P, df_S)  
+    sheets_dict = write_excel_model(model, f'{args.outdir}/{model.id}.mkmodel.xlsx', args.nofigs, None, df_B, df_P, df_S)  
     logger.info(f"'{args.outdir}/{model.id}.mkmodel.xlsx' created!")