FIlter data and bypass broken loadData conditions argument

The "conditions" argument for loadData in SSIT.m is not properly filtering by Dex concentration. Have pre-processed data with Python for now.

Author: alexpopinga

WorkSpace/EricModel/EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_1.csv

@@ -0,0 +1,34 @@
+import pandas as pd
+import numpy as np
+
+# Load the CSV file
+file_path = 'pdoCalibrationData_NeuertLab_240925.csv'
+data = pd.read_csv(file_path)
+
+# Filter the data based on Condition = '2M_NaCl_Step'
+filtered_data = data.loc[data['Condition'].str.contains('2M_NaCl_Step', case=False, na=False), 
+                         ['Cell_id', 'Condition', 'Replica', 'Conc_mM', 'Time_index_min', 
+                          'RNA_STL1_cyto_TS3Full', 'RNA_STL1_nuc_TS3Full', 'Cyto_STL1_avg_int_TS3Full', 'Nuc_STL1_avg_int_TS3Full']]
+
+# Sum 'RNA_STL1_cyto_TS3Full' and 'RNA_STL1_nuc_TS3Full' into new column 'RNA_STL1_total_TS3Full'
+filtered_data['RNA_STL1_total_TS3Full'] = filtered_data['RNA_STL1_cyto_TS3Full'] + filtered_data['RNA_STL1_nuc_TS3Full']
+
+# Sum 'Cyto_STL1_avg_int_TS3Full' and 'Nuc_STL1_avg_int_TS3Full' into new column 'STL1_avg_int_TS3Full' and transform it into an integer for SSIT
+#filtered_data['STL1_avg_int_TS3Full'] = (filtered_data['Cyto_STL1_avg_int_TS3Full'] + filtered_data['Nuc_STL1_avg_int_TS3Full']).astype(int)
+# Fill NaN or inf values with 0, then sum and convert to integer
+filtered_data['STL1_avg_int_TS3Full'] = (
+    filtered_data['Cyto_STL1_avg_int_TS3Full'].replace([np.inf, -np.inf], np.nan).fillna(0) +
+    filtered_data['Nuc_STL1_avg_int_TS3Full'].replace([np.inf, -np.inf], np.nan).fillna(0)
+).astype(int)
+
+
+# Replace 'Time_index_min' with 'time' to be read by SSIT.m
+filtered_data.rename(columns={'Time_index_min': 'time'}, inplace=True)
+
+# Define the output file path
+output_file = 'filtered_data_2M_NaCl_Step.csv'
+
+# Save the filtered data to a new CSV file
+filtered_data.to_csv(output_file, index=False)
+
+output_file

WorkSpace/EricModel/EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_10.csv

@@ -0,0 +1,29 @@
+import pandas as pd
+
+# Load the CSV file
+file_path = "GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03.csv"
+df = pd.read_csv(file_path)
+
+# Check the first few rows to understand the structure
+df.head()
+
+# Define the required columns
+required_columns = ["nucGRint", "cytoGRint", "time", "dex_conc", "unique_cell_id", "normGRcyt", "normGRnuc"]
+
+# Filter the DataFrame to only include the required columns
+df_filtered = df[required_columns]
+
+# Define the output file paths
+output_files = {
+    1: "GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_1.csv",
+    10: "GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_10.csv",
+    100: "GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_100.csv"
+}
+
+# Filter and save the data for each Dex_Conc value
+for conc, path in output_files.items():
+    df_filtered_conc = df_filtered[df_filtered["dex_conc"] == conc]
+    df_filtered_conc.to_csv(path, index=False)
+
+# Provide confirmation
+output_files

WorkSpace/EricModel/EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_100.csv

@@ -58,7 +58,7 @@
 else
     %% STEP 0.B. -- Create Base Model for GR Only
     %    Here, I set up the model for the GR translocation dynamics.
-    fitOptions = optimset('Display','iter','MaxIter',300);
+    fitOptions = optimset('Display','iter','MaxIter',200);
 
     % Create blank SSIT model.
     ModelGR = SSIT;
@@ -115,10 +115,28 @@
     ModelGRparameterMap = cell(1,size(GRfitCases,1));
     ModelGRfit = cell(1,size(GRfitCases,1));
     % ModelGRODEfit = cell(1,size(GRfitCases,1));
+
+    % OLD GR DATA
+    % ModelGRfit{1} = ModelGR.loadData("EricData/Gated_dataframe_Ron_020224_NormalizedGR_bins_Dex_Conc_1.csv",...
+    %           {'nucGR','normgrnuc';'cytGR','normgrcyt'});
+    % ModelGRfit{2} = ModelGR.loadData("EricData/Gated_dataframe_Ron_020224_NormalizedGR_bins_Dex_Conc_10.csv",...
+    %           {'nucGR','normgrnuc';'cytGR','normgrcyt'});
+    % ModelGRfit{3} = ModelGR.loadData("EricData/Gated_dataframe_Ron_020224_NormalizedGR_bins_Dex_Conc_100.csv",...
+    %           {'nucGR','normgrnuc';'cytGR','normgrcyt'});
+
+    % NEW GR DATA
+    ModelGRfit{1} = ModelGR.loadData("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_1.csv",...
+              {'nucGR','normGRnuc';'cytGR','normGRcyt'});
+    ModelGRfit{2} = ModelGR.loadData("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_10.csv",...
+              {'nucGR','normGRnuc';'cytGR','normGRcyt'});
+    ModelGRfit{3} = ModelGR.loadData("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03_dex_conc_100.csv",...
+              {'nucGR','normGRnuc';'cytGR','normGRcyt'});
+    
     for i=1:3
-        ModelGRfit{i} = ModelGR.loadData("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03.csv",... %ModelGRfit{i} = ModelGR.loadData("EricData/Gated_dataframe_Ron_020224_NormalizedGR_bins.csv",... %
-            {'nucGR','normGRnuc';'cytGR','normGRcyt'},...
-            {'Dex_Conc',GRfitCases{i,2}});
+        % ModelGRfit{i} = ModelGR.loadData("EricData/Gated_dataframe_Ron_020224_NormalizedGR_bins.csv",...
+        %     {'nucGR','normgrnuc';'cytGR','normgrcyt'},...
+        %     {'Dex_Conc',GRfitCases{i,2}});        
+        %ModelGRfit{i} = ModelGR.loadData("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03.csv",{'nucGR','normGRnuc';'cytGR','normGRcyt'},
         ModelGRfit{i}.parameters(13,:) = {'Dex0', str2num(GRfitCases{i,1})};
         ModelGRparameterMap(i) = {(1:8)};
         % parameters 1 - 8 refer to the parameter set that is relevant to
@@ -159,12 +177,11 @@
 
     % STEP 1.C. -- Combine all three GR models and fit using a single parameter set.
     for jj = 1:fitIters
-        combinedGRModel = SSITMultiModel(ModelGRfit,ModelGRparameterMap,logPriorGR);
-        combinedGRModel = combinedGRModel.initializeStateSpaces(boundGuesses);
-        combinedGRModel = combinedGRModel.updateModels(GRpars,false);
-        GRpars = combinedGRModel.maximizeLikelihood(...
-            GRpars, fitOptions);
-        save('EricModel_MMDex','GRpars') 
+         combinedGRModel = SSITMultiModel(ModelGRfit,ModelGRparameterMap,logPriorGR);
+         combinedGRModel = combinedGRModel.initializeStateSpaces(boundGuesses);
+         combinedGRModel = combinedGRModel.updateModels(GRpars,false);
+         GRpars = combinedGRModel.maximizeLikelihood(GRpars, fitOptions);
+         save('EricModel_MMDex','GRpars') 
     end
 
     save('EricModelGR_MMDex','GRpars','combinedGRModel', 'ModelGRfit', 'log10PriorStd') 
@@ -213,16 +230,39 @@
 end 
 
 %%
-for i=1:3
-    combinedGRModel.SSITModels{i} = combinedGRModel.SSITModels{i}.loadData("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03.csv",...
-                {'nucGR','normGRnuc';'cytGR','normGRcyt'});
-end
+% dex_conc_values = [1, 10, 100];
+% 
+% for i = 1:3
+%     % Load the full dataset first
+%     fullData = readtable("EricData/GR_ALL_gated_with_CytoArea_and_normGR_Feb2825_03.csv");
+% 
+%     % Filter based on dex_conc value
+%     filteredData = fullData(fullData.dex_conc == dex_conc_values(i), :);
+% 
+%     % Save filtered data to a temporary file (optional if loadData can take a table directly)
+%     tempFile = sprintf("temp_filtered_data_%d.csv", dex_conc_values(i));
+%     writetable(filteredData, tempFile);
+% 
+%     % Load the filtered data
+%     combinedGRModel.SSITModels{i} = combinedGRModel.SSITModels{i}.loadData(tempFile, ...
+%                 {'nucGR','normGRnuc';'cytGR','normGRcyt'});
+% end
+
 
 %%     STEP 1.F. -- Make Plots of GR Fit Results
 makeGRPlots(combinedGRModel,GRpars)
 
+
+%combinedGRModel = combinedGRModel.updateModels(GRpars,false);
+%for i=1:length(ModelGRfit)
+    %  Update parameters in original models.
+%    ModelGRfit{i} = combinedGRModel.SSITModels{i};
+%    ModelGRfit{i}.tSpan = sort(unique([ModelGRfit{i}.tSpan,linspace(0,180,30)]));
+%    ModelGRfit{i}.makeFitPlot([],1,[],true,'STD')
+%end
+
 save('EricModelGR_MMDex','GRpars','combinedGRModel','MHResultsGR') 
-save('workspaceMay4_2024.mat','GRpars', 'ModelGRfit', 'combinedGRModel','MHResultsGR', 'log10PriorStd', 'ModelGRparameterMap')
+save('workspaceMay4_2025.mat','GRpars', 'ModelGRfit', 'combinedGRModel','MHResultsGR', 'log10PriorStd', 'ModelGRparameterMap')
 
 %%  STEP 2 -- Extend Model to Include DUSP1 Activation, Production, and Degradation
 if loadPrevious

WorkSpace/EricModel/EricData/Gated_dataframe_Ron_020224_NormalizedGR_bins_Dex_Conc_1.csv

@@ -214,7 +214,10 @@ function makeSeparatePlotOfData(app,smoothWindow,fignums,usePanels,varianceType,
 cols = ['b','r','g','m','c','k'];
 cols2 = [.90 .90  1.00; 1.00 .90 .90; .90 1.00 .90; .60 .60  1.00; 1.00 .60 .60; .60 1.00 .60];
 LG = {};
+
+%% 1:NdDat can be changed to plot species separately (e.g., separate cytGR and nucGR plots)
 for iplt=1:NdDat
+%for iplt=2:2
     % if Plts_to_make(iplt)
     switch varianceType
         case 'STD'
@@ -235,7 +238,9 @@ function makeSeparatePlotOfData(app,smoothWindow,fignums,usePanels,varianceType,
 
 %% Add data to trajectory plot.
 for iTime = 1:numTimes
+    %% 1:NdDat can be changed to plot species separately (e.g., separate cytGR and nucGR plots)
     for j=1:NdDat
+    %for j=2:2 
         mnsDat(iTime,j) = [0:length(dataHistTime{iTime,j})-1]*dataHistTime{iTime,j};
         mns2Dat(iTime,j) = [0:length(dataHistTime{iTime,j})-1].^2*dataHistTime{iTime,j};
         % Compute the 25% and 75% range from data.
@@ -254,7 +259,9 @@ function makeSeparatePlotOfData(app,smoothWindow,fignums,usePanels,varianceType,
 end
 varDat = mns2Dat-mnsDat.^2;
 T_array = app.DataLoadingAndFittingTabOutputs.fittingOptions.dataTimes;
+%% 1:NdDat can be changed to plot species separately (e.g., separate cytGR and nucGR plots)
 for j=1:NdDat
+%for j=2:2
     switch varianceType
         case 'STD'
             errorbar(meanVarTrajAxis,T_array,...